studying an all-vanadium photoelectrochemical cell …
TRANSCRIPT
STUDYING AN ALL-VANADIUM PHOTOELECTROCHEMICAL CELL FOR HIGHLY
EFFICIENT SOLAR ENERGY CONVERSION AND STORAGE VIA
PHOTOCATALYTIC REDOX REACTION
by
DONG LIU
Presented to the Faculty of the Graduate School of
The University of Texas at Arlington in Partial Fulfillment
of the Requirements
for the Degree of
DOCTOR OF PHILOSOPHY
THE UNIVERSITY OF TEXAS AT ARLINGTON
AUGUST 2015
ii
Copyright © by Dong Liu 2015
All Rights Reserved
iii
Acknowledgements
They all say that this is one of the most important sections in one's PhD thesis, I
could not agree more as this is where readers get to know the author personally.
Except those hardships I had and sacrifice I made in last five years, the state of
my mind toward the future life through this PhD training is way more influential than the
training itself. For that, I am very very grateful! To this end, I like to express my sincere
thanks firstly to my supervisor Dr. Fuqiang Liu. I would not be able to complete this
program without his continuous encouragement, knowledge, technical help and financial
support in last five years. Not only did he work with me closely on things we have little
knowledge or experience of from the scratch, but also he taught me ways to conduct
carefully designed, analytically meticulous and logically rigorous research. Most
important, he is the only person who let me feel that I am actually doing something great
for the first time in my whole academic life combined. It is such a privilege to ever be one
of his students!
Secondly, I would like to thank my committee members Dr. Krishnan Rajeshwar,
Dr. Efstathios "Stathis" Meletis, Dr. Weidong Zhou, Dr. Yaowu Hao, Dr. Kyungsuk Yum,
and Dr. Harry F. Tibbals for their valuable comments, suggestions and criticisms during
my comprehensive and dissertation defense. My parents always told me to be grateful for
those who sound harsh during my study, so special thanks go to Dr. Rajeshwar for his
critical but beneficial criticism to help me stay on the right track of this research.
In addition, I sincerely appreciate the scientific discussion, perspective, criticism,
technical aid and mental support from all of my past and current lab mates, Dr. Noor
Siddique, Dr. Chiajen Hsu, Dr. Mingsheng Wei, Wei Zi, Syed Dawar Sajjad, Yi Shen,
Snigdha Rajendra Rashinkar, Shambhavi Raju Sakri, Amir Hosein Salehi Gilani, Sonam
Patel for my research through which we made very good friends and colleagues. They
iv
created a very friendly and professional work environment in our lab. Among them, Zi
worked with me on this research very closely and her contribution to this research is one
of the reason I can go this far. She and Dr. Noor Siddique are much more than just a
colleague for their kindness, patience, endurance, encouragement and trust to me all the
time!
Furthermore, I also need to express the gratitude to the Characterization Center
for Materials and Biology (CCMB) in my department for providing me instrument
trainings. Since vast characterization work needs to be done in the research, I would not
imagine this possible without this well-maintained center and its facilities by the lab
manager Dr. Jiechao Jiang and lab technician David Yan.
All faculty members from Department of Materials Science and Engineering have
taught me comprehensive and useful courses within the program, I say thank you to all of
them by heart regardless of my future professional career pathway. All past and current
department staffs, Beth Robinson, Jennifer Standlee and Libia Cuauhtli are also praised
for offering me generous help.
Meanwhile, my friends, Manouchehr Teimouri, Jain, Astha, Vibhu Sharma, Kush
Shah, Felipe Monte, Ami, Shah, Gaurav Nagalia, Akshay Hande, Orathai Thumthan,
Randle Kelton, Derek Wong, Yishu Wang, Minghui Zhang, Jacky Lu, Kunhua Yu, Aaron
Chiu, Ruiqian Jiang, Yang Li, Golsa Mortazavi, Vinay Sharma, Sunil Sahi, Cheng Sheng,
Shuye Wang, Wenbo Zhu, Fuqiang Zhang, Hongzhu Liao, Yuetao Sun, Cancan Xu,
Xiaosong Duan, Dr. Fangcheng Xu, Dr. Lei Qiao, Dr. Xueyang Cheng, Dr. Chengdong
Xu, Dr. Nan Zhang, Dr. Wei Li, Dr. Jun Zhou, Dr. Ye Zhou and their company in last five
years are appreciated!
Last but not the least, I am indefinitely indebted to all my family. My wife is the
most beautiful, elegant, considerate and amazing woman I have ever seen. Her endless
v
faith, support, inspiration, trust and love for me constitutes the cornerstone behind this
degree. My son is one of the greatest gift I have received from HIM and his birth
completes this family. My parents especially my mother upheld their son since he was 18
to travel thousands of miles within the country and even across the ocean to pursue his
dream, I cannot even begin to imagine the enormous sacrifice, faith, trust and love they
had for me. For that, I am grateful forever! My parents-in-law and their great
encouragement, support and love are also one of the reasons this challenge becomes
much easier.
I apologize to those who accompanied and helped me and whose name I forgot
to mention in this acknowledgement.
July 14, 2015
vi
Abstract
STUDYING AN ALL-VANADIUM PHOTOELECTROCHEMICAL CELL FOR HIGHLY
EFFICIENT SOLAR ENERGY CONVERSION AND STORAGE VIA
PHOTOCATALYTIC REDOX REACTION
Dong Liu, PhD
The University of Texas at Arlington, 2015
Supervising Professor: Fuqiang Liu
This study aims at developing a promising alternative approach to utilize solar
energy continuously with high efficiency by employing photocatalytic vanadium redox
reactions to overcome major drawbacks associated with conventional
photoelectrochemical water splitting.
The preliminary study was implemented on TiO2 and WO3/TiO2 by linear sweep
voltammetry (LSV), cyclic voltammetry (CV) and zero-resistance ammetry (ZRA) in a
photoelectrochemical cell (PEC). The results show much enhanced photoelectrochemical
response with the assistance of vanadium(IV, VO2+
) redox species. Such photoresponse
improvement is attributed to the hole scavenging effect by fast vanadium redox reaction
to depress charge recombination at semiconductor/liquid interface in both TiO2 and
WO3/TiO2 cases. Facilitated electron transfer from WO3 to TiO2 further contributes to the
photoresponse of WO3/TiO2 hybrid electrode due to their band structures interplay.
During the study, a unique photocharging/discharging process was discovered
and modeled on WO3-based hybrid electrodes due to the formation/decomposition of
hydrogen tungsten bronze (HxWO3) from WO3 under AM 1.5 illumination in highly acidic
environment. Comparison of electrochemical impedance spectroscopy (EIS) results on
vii
TiO2 and WO3/TiO2 electrodes indicates that vanadium redox species, due to their fast
electrochemical kinetics, not only enhance the photocurrents under illumination, but also
help release reversely the stored electrons under dark through decomposition reaction of
HxWO3.
To further improve photocatalytic property of the PEC, the conventional H2SO4
was replaced with a novel supporting electrolyte, methanesulfonic acid (MSA) as a
simple but feasible optimization measure. CV, zero-resistance ammetry (ZRA) and
conductivity measurements demonstrate its excellent chemical stability, high conductivity
and superior photoelectrochemical performance over H2SO4, especially when vanadium
redox species were involved in the electrolyte. The root cause of such ability of MSA
relies on its prolonged electron lifetime and being capable of dramatically diminishing
charge transfer resistance and interfacial capacitance at photoelectrode/electrolyte
interface under AM1.5 illumination, according to EIS Nyquist plot and Bode plot results.
To quantify the efficiency of proposed all-vanadium PEC, the Faradaic efficiency
and Incident photon-to-electron conversion efficiency of the cell were measured. While
the Faradaic efficiency of the cell reaches to 84.8% after a 60-h cell operation, the IPCE
of the cell was improved on V-MSA by a factor of 18.6, 9.7 and 2.5 compared to pure
H2SO4, V-H2SO4 and pure MSA respectively. Such remarkable IPCE enhancement of the
cell is believed due to synergy between fast vanadium redox kinetics and prolonged
electron life time of MSA.
viii
Table of Contents
Acknowledgements ............................................................................................................. iii
Abstract .............................................................................................................................. vi
List of Illustrations ............................................................................................................... x
List of Tables ..................................................................................................................... xv
Nomenclature ....................................................................................................................xvi
CHAPTER 1 Introduction .................................................................................................... 1
Chapter 2 Background Information and Objective of the Study ......................................... 6
2.1 Introduction ............................................................................................................... 6
2.2 Operating Principles of the PEC ............................................................................... 6
2.3 Photoelectrochemical Water Splitting ....................................................................... 7
2.3.1 Thermodynamic Requirement ........................................................................... 8
2.3.2 Kinetic Requirement .......................................................................................... 9
2.4 Photoelectrochemical Solar Energy Storage .......................................................... 10
2.4.1 Redox-assisted Solar Energy Conversion and Storage .................................. 15
2.5 Scope of Study and Objectives .............................................................................. 17
2.6 Description of This Dissertation .............................................................................. 20
Chapter 3 Primary Study of All-Vanadium PEC Using Different
Photoelectrodes ................................................................................................................ 22
3.1 Introduction ............................................................................................................. 22
3.2 Experiments ............................................................................................................ 22
3.2.1 Electrode and Electrolyte Preparation ............................................................. 22
3.2.2 Cell Design and Assembly .............................................................................. 23
3.2.3 Material Characterization ................................................................................ 24
3.3 Photoelectrochemical Study ................................................................................... 31
ix
3.3.1 Proof-of-the-concept Experiment .................................................................... 31
3.3.2 Photoelectrochemical Performance of the Hybrid Electrode .......................... 34
3.3.3 Preliminary Electron Storage Study by Hybrid Electrode ................................ 38
3.4 Conclusion .............................................................................................................. 41
Chapter 4 Electron Storage Study of Hybrid Electrode .................................................... 44
4.1. Introduction ............................................................................................................ 44
4.2 Experiments ............................................................................................................ 44
4.3 Zero-resistance Ammetry Study ............................................................................. 49
4.4 Electrochemical Impedance Spectroscopy Study .................................................. 57
4.5 Electron Storage Mechanism ................................................................................. 63
4.6 Conclusion .............................................................................................................. 66
Chapter 5 Electrolyte Study of All-Vanadium PEC ........................................................... 68
5.1 Introduction ............................................................................................................. 68
5.2 Experiments ............................................................................................................ 69
5.3 Cyclic Voltammetry and Zero-resistance Ammetry Study ...................................... 71
5.4 Bulk Ionic Conductivity Study ................................................................................. 76
5.4 Electrochemical Impedance Spectroscopy Study .................................................. 77
5.5 Long-term Stability Study ....................................................................................... 81
5.6 Efficiencies Measurement ...................................................................................... 83
5.7 Conclusion .............................................................................................................. 87
Chapter 6 Conclusion ........................................................................................................ 89
Appendix Achievements.................................................................................................... 92
References ........................................................................................................................ 96
Biographical Information ................................................................................................. 104
x
List of Illustrations
Figure 1-1 Three approaches known for storing solar energy: Natural photosynthesis
(left), PEC (middle), and PV (right). .................................................................................... 2
Figure 2-1 Operation principle of photoelectrochemical cells based on n-type
semiconductor. a) Regenerative-type cell producing electric current from sunlight; b) A
cell that generates a chemical fuel, hydrogen, through the photo-cleavage of water. ....... 7
Figure 2-2 Band positions of common semiconductor materials in contact with aqueous
electrolyte at pH=1. Two red dashed lines indicate hydrogen and oxygen evolution
reaction potential. Several redox couple standard reduction potentials are also
represented against NHE level. .......................................................................................... 8
Figure 2-3 Schematic of a photoelectrochemical solar cell combining both solar
conversion and storage capabilities. (a) Under illumination; (b) Under dark. ................... 11
Figure 2-4 Cell structure of a PESC with (left) insoluble redox couple and (right) soluble
redox couple. ..................................................................................................................... 12
Figure 2-5 Cell structure of a PESC with a third electrode (counter electrode) in the
photoelectrode compartment. P: Photoelectrode; A: Counter electrode; M: Membrane; S:
Storage electrode; EL: Electrolyte; F: Electrical Switch and L: Load. ............................... 14
Figure 2-6 Schematic of the photocatalytic generation of H2 under the proposed hole
shuttle mechanism. ........................................................................................................... 16
Figure 2-7 Kinetics comparison of reduction reaction of O2/H2O and VO2+
/VO+ in 3 M
H2SO4. ............................................................................................................................... 18
Figure 2-8 Schematic of the solar energy storage approach via photocatalytic redox
reactions. ........................................................................................................................... 19
Figure 3-1 The representative schematics of (a) half-cell configuration and (b) full-cell
configuration. ..................................................................................................................... 24
xi
Figure 3-2 XRD spectra of the TiO2 and WO3/TiO2 (12 wt% WO3) hybrid electrodes. ..... 25
Figure 3-3 Raman spectra of the TiO2 and WO3/TiO2 (12 wt% WO3) hybrid electrodes. . 26
Figure 3-4 SEM images of sintered TiO2 sample in (A) lower and (B) higher magnification,
together with (C) unsintered TiO2 powder. ........................................................................ 27
Figure 3-5 SEM image of the WO3/TiO2 (12 wt% WO3) hybrid electrodes. A higher-
magnification image is shown in the inset......................................................................... 28
Figure 3-6 EDS mapping result of the WO3/TiO2 (12 wt% WO3) hybrid electrode. .......... 29
Figure 3-7 UV-Vis diffuse reflectance spectra of the TiO2 and WO3/TiO2 (12 wt% WO3)
hybrid electrodes. .............................................................................................................. 30
Figure 3-8 The linear sweep voltammograms of the TiO2 electrode under various
electrochemical conditions in a half-cell configuration. The scan rate was 5 mV/s. ......... 31
Figure 3-9 The plot of (a) IvAM1.5/I
wAM1.5 and (b) I
vAM1.5/I
vDark as a function of vanadium
species concentration at 0.5 V and 1V. ............................................................................ 34
Figure 3-10 Linear sweep voltammograms of three different electrodes in 3 M H2SO4
electrolyte in a half-cell configuration under alternate dark and illumination conditions.
The scan rate was 5 mV/s. ............................................................................................... 35
Figure 3-11 Photocurrents of different photoelectrodes using ZRA method in (A) 3 M
H2SO4 electrolyte and (B) various vanadium electrolytes in different cell configurations.
The labels in Figure 3-11B designate: 1: TiO2 in 0.01 M V(IV)-H2SO4, half-cell; 2: TiO2 in
0.01 M V-H2SO4, full-cell; 3: Hybrid in 0.01 M V(IV)-H2SO4, half-cell; and 4: Hybrid in 0.01
M V-H2SO4, full-cell. .......................................................................................................... 37
Figure 3-12 CV of the WO3 electrode in 3 M H2SO4 electrolyte under dark and AM 1.5
illumination in a half-cell configuration. The scan rate was 20 mV/s. ............................... 40
Figure 4-1 XRD spectra of (a) TiO2 electrode, and hybrid electrodes with various WO3
loadings: (b) 6 wt%, (c) 12 wt%, and (d) 24 wt%. ............................................................. 46
xii
Figure 4-2 SEM images of (A) TiO2 electrode and hybrid electrodes with different WO3
loadings (B) 6 wt%, (C) 12 wt%, and (D) 24 wt%. High-magnification images are shown in
the inset for each figure. The scale bars in low-magnification and high-magnification
images represent 300 um and 10 um respectively. .......................................................... 47
Figure 4-3 Raman Spectra of the hybrid electrode (24 wt% WO3) before and after ZRA
experiments in 3 M H2SO4 electrolyte. .............................................................................. 48
Figure 4-4 Photocurrent measurements of PECs with different electrodes in 3 M H2SO4
electrolytes under alternative dark/illumination conditions. The inset graph shows the
photocurrent of different electrodes from 5 to 10 min under illumination. ......................... 50
Figure 4-5 Appearance of hybrid WO3/TiO2 electrodes with WO3 loadings of (A) 1 wt%,
(B) 6 wt%, (C) 12 wt%, and (D) 24 wt% before and after ZRA experiment in 3 M H2SO4
electrolytes. ....................................................................................................................... 51
Figure 4-6 XRD spectra of the hybrid electrode (24 wt% WO3) in the range of (a) 20 – 80°
(2θ) and (b) 22.5 - 25° (2θ) showing WO3 characteristic peaks before and after ZRA
experiments in 3 M pure acid electrolyte. ......................................................................... 53
Figure 4-7 Photocurrent measurements of PECs with different electrodes in 0.01 M all-
vanadium electrolytes under alternative dark/illumination conditions. .............................. 54
Figure 4-8 Significant electron storage capacity of the hybrid electrode (24 wt% WO3)
compared to the TiO2 electrode in 0.01 M all-vanadium electrolytes after the first 5 min
illumination. The graph was rearranged from Figure 4-7. ................................................. 56
Figure 4-9 Nyquist plots of PECs in (A) 3 M H2SO4 and (B) 0.01 M all-vanadium with a
TiO2 photoelectrode. The inset graph in Figure 4-9B represents Nyquist plots of the cell in
all tested frequency. .......................................................................................................... 58
Figure 4-10 Nyquist impedance spectra of PECs in (A) 3 M H2SO4 electrolytes and (B)
0.01 M all-vanadium electrolytes with a hybrid (24 wt% WO3) photoelectrode. The inset
xiii
graph in Figure 4-10B represents Nyquist plots of the cell under various conditions in all
tested frequencies. ............................................................................................................ 60
Figure 4-11 Appearance of hybrid WO3/TiO2 electrodes with WO3 loadings of (A) 1 wt%,
(B) 6 wt%, (C) 12 wt%, and (D) 24 wt% before and after ZRA experiment in 0.01 M all-
vanadium electrolytes. ...................................................................................................... 62
Figure 4-12 Schematic representation of proposed electron storage mechanism and
charge transfer pathways for the hybrid WO3/TiO2 electrode in the all-vanadium
photoelectrochemical storage cell. .................................................................................... 63
Figure 5-1 Cyclic voltammetry of a PEC using pure MSA as the electrolyte and TiO2 as
the photoelectrode with scan range of -0.5 to 2.1V under dark and AM 1.5 illumination.
The scan rate was 20 mV/s. ............................................................................................. 72
Figure 5-2 Linear sweep voltammetry of a PEC using TiO2 as the photoelectrode in 3 M
H2SO4 and 3 M MSA electrolytes under dark and AM 1.5 illumination. ............................ 73
Figure 5-3 Photocurrents of TiO2 photoelectrode using ZRA method in various
electrolytes under alternate dark and AM 1.5 illumination conditions. .............................. 75
Figure 5-4 Nyquist plots of the test cell using various electrolytes. .................................. 76
Figure 5-5 Electrochemical impedance spectroscopy of a PEC in the form of (A) Nyquist
plot and (B) Bode plot using TiO2 as the photoelectrode in various electrolytes under AM
1.5 illumination. ................................................................................................................. 78
Figure 5-6 Long-term photoelectrochemical profile of all-vanadium PEC using the TiO2
photoelectrode immersed in 0.01 M V-3 M MSA electrolytes under AM 1.5 illumination for
60 hrs. ............................................................................................................................... 82
Figure 5-7 XRD spectra of TiO2 photoelectrode before and after 60 hrs cell operation. .. 83
Figure 5-8 UV-vis absorbance spectra of 0.01 M vanadium(IV) ions in 3 M MSA
electrolyte at different period of time of photocharging. .................................................... 84
xiv
Figure 5-9 IPCE of the PEC using a TiO2 photoelectrode in various electrolytes. ........... 86
xv
List of Tables
Table 1-1 Comparison between semiconductor-liquid junction photoelectrochemical cells
with solid state junction solar cells. ..................................................................................... 3
Table 2-1 List of exemplary photoelectrochemical conversion and storage modes using a
two-electrode configuration. .............................................................................................. 12
Table 5-1 Calculated ionic conductivity of various electrolytes......................................... 77
Table 5-2 The maximum frequency of the peak and calculated electron lifetime of a TiO2
photoelectrode in various electrolytes under AM 1.5 illumination. .................................... 81
xvi
Nomenclature
Abbreviation Meaning
PEC photoelectrochemcical cell
PV photovoltaic
PESC photoelectrochemical storage cell
SCR space charge region
HER hydrogen evolution reaction
OER oxygen evolution reaction
STH solar-to-hydrogen
VRB vanadium redox battery
CVD chemical vapor deposition
PLD pulsed laser deposition
PVD physical vapor deposition
CB conduction band
VB valence band
LSV linear cyclic voltammetry
CV cyclic voltammetry
ZRA zero-resistance ammetry
EIS electrochemical impedance spectroscopy
WE working electrode
CE counter electrode
RE reference electrode
NHE normal hydrogen electrode
XRD x-ray diffraction
SEM scanning electron microscope
EDS Energy-dispersive x-ray spectroscopy
JCPDS joint committee on powder diffraction standards
MSA methanesulfonic acid
IPCE incident photon-to-current conversion efficiency
1
CHAPTER 1
Introduction
Solar energy, including radiant light and heat from the sun, has been harnessed
by humans since ancient times using a range of ever-evolving technologies. These
technologies include indirect methods using meteorological or geophysical effects, or
direct methods as use of solar energy in the form of heat, generation of electric power by
photovoltaic process and storage the solar energy by converting it to chemical fuels.[1, 2]
The indisputable fact that more energy from sunlight strikes earth in 1 hour than all the
energy consumed by humans in an entire year is the sole driving force for mankind to
promote wider and deeper utilization of solar energy.[2, 3]
Among those generic methods mentioned above to utilize solar energy, three
approaches, natural photosynthesis, photoelectrochemical cell (PEC) and photovoltaic
(PV) cell are known for producing storable energy from sunlight as demonstrated in
Figure 1-1.[4] Natural photosynthesis in plants relies on solar energy to convert carbon
dioxide (CO2) and water (H2O) to oxygen (O2) and glucose (a source of fuel) with a
conversion efficiency of about 3%. The PEC transforms the sun's energy into an electrical
current, or a chemical fuel, such as hydrogen (H2), from a liquid such as water. The most
effective PECs can produce chemical fuels with a conversion efficiency of about 10% and
electricity with an efficiency of about 16% while PV cells convert light to electricity with an
efficiency as high as 25% even back in 90’s.[4] With its high energy conversion efficiency
and the development of semiconductors, mainly Si, photovoltaic cell or solar cell, as a
potentially widespread approach to sustainable energy production, is receiving
heightened attention. This is because increased economical, environmental and
ecological concerns toward carbon-neutral energy production have been seriously taken
by human beings. So far, the solar cells have been the dominating devices in which the
2
junction is formed between inorganic solid-state materials, usually doped forms of
crystalline or amorphous silicon, which development has benefited from the experience
and material availability in the semiconductor industry.[5]
Figure 1-1 Three approaches known for storing solar energy: Natural photosynthesis
(left), PEC (middle), and PV (right).
However, the inorganic solid-state junction solar cell faces new challenges in the
coming years mainly due to material-to-device manufacturing cost.[2] For example, the
sale price of grid-connected PV electricity must excel a $0.25-0.30 per kilowatt-hour
(kWh) threshold to recover the initial capital investment and cost over the lifetime of the
PV operation for a practical PV module with efficiency of 15-20%. The demand for
producing cost-effective PV module is crucial in every level. Another key issue hurdling
scalable implementation of PV technology lies in the trade-off between material purity
and device performance.[2] In a typical planar solar cell design, the charge carriers are
collected in the same direction as light is absorbed. A minimum thickness of the cell is set
by the thickness of material required to absorb >90% of the incident sunlight. However,
the required thickness of the material also imposes a constraint on the required purity of
the material, because the photogenerated charge carriers must live sufficiently long
enough within the absorbing material to arrive at the electrical junction, where they can
be separated to produce an electrical current flow through the metallic contacts to the
3
cell. In other words, impure absorber materials with short charge carrier lifetimes can
effectively absorb sunlight but cannot convert it effectively. On the contrary, absorber
materials with the necessary purity are generally costly to produce and manufacture. This
cost-thickness-purity constraint, combined with expensive manufacturing cost is largely
the cause why all current PV cells are incomparable to utility-scale power generation
approaches.[2, 3]
Table 1-1 Comparison between semiconductor-liquid junction photoelectrochemical cells
with solid state junction solar cells.
Due to the continuing, rapid worldwide progress in nano-science/technology and
photoelectrochemistry, a new category of solar energy utilization, photoelectrochemical
cell, has been invented featuring a solid-liquid junction formed between the
semiconductor and the electrolyte interface. In contrast to solid state photovoltaic
devices, semiconductor/liquid junction based photoelectrochemical cells are relatively
4
inexpensive involving convenient and economically viable material synthesis and tailoring
processes, and are easy to fabricate. The semiconductor/liquid configuration offers the
following four advantages:[3] (1) the junction required for efficient charge separation of
photogenerated electrons and holes is very easily formed by simply immersing the
semiconductor in an appropriate electrolyte solution; (2) the liquid electrolyte offers the
capability of a readily conformable and strain-free junction; (3) a third electrode can be
added to PEC cells to provide in-situ chemical storage for 24 hr/day power; and (4) the
conversion of solar energy directly into fuel eliminates the need for external wires and a
separate electrolyzer. The PEC approach to solar energy conversion has achieved high
efficiencies for both electrical power (>15%) and hydrogen generation (>10%) to date.
Compared to solid-state junction solar cells, the most important and vital advantage given
in Table 1-1 in PECs is the possibility of in situ storage[3, 6] because PECs naturally offer
the opportunity to integrate the energy conversion and storage functions.
PEC has been shown to directly split water into H2 and O2, thereby providing a
promising basis for the renewable, clean production of hydrogen from sunlight. The
materials can be cheap polycrystalline forms, because of the relaxed requirements on the
minority carrier diffusion length. However, the known materials, which are robust in water
splitting, are not responsive to a wide portion of the solar radiation spectrum; they work
best in the ultraviolet (UV) region- yielding relatively low solar-to-electrical (STE)
efficiencies. Finding new photoelectrodes, either individually or in combination, that can
allow the efficient, integrated conversion of sunlight to chemical fuels, is one of the
primary aims of solar energy conversion research. The goal is to identify PEC systems
that display the same efficiency and stability for visible-light-induced water splitting as
those demonstrated for near-UV-light-induced water splitting.[3] Closing this gap will lead
to the development of cheap and efficient systems that, in an integrated fashion, could
5
produce chemical fuels (e.g., H2) directly from sunlight and therefore directly address, not
only the conversion, but also the storage issues, associated with solar energy conversion
schemes.
This Ph.D. dissertation primarily focused on the three major challenges that have
hindered the development of efficient solar energy storage using PECs in the past, i.e.,
continuous energy conversion (despite the intermittent nature of sunlight), high capacity
energy storage medium, and reversible conversion reaction kinetics. This research
strives to establish a methodology for the systematic design and effective integration of
high efficiency regenerative solar energy storage using all vanadium redox species, as an
alternative to photoproduction of H2. Most important, a tandem WO3/TiO2photoelectrode
was designed to effectively tune the photogenerated charge flow, so that the electrons
could be stored in the photoelectrode upon solar illumination and released under dark,
allowing potentially uninterrupted solar energy conversion.
6
Chapter 2
Background Information and Objective of the Study
2.1 Introduction
The discovery by Fujishima and Honda[7] in 1972 offered the prospect of using
sunlight to produce environmentally-friendly fuel. However, the critical hurdle preventing
its wide-spread implementation is hydrogen handling. Although molecular hydrogen has
very high energy density on a mass basis, as a gas at ambient conditions it has very low
energy density by volume. Pressurization and liquefaction increase volumetric density,
but also bring inconvenience and safety concerns. Moreover, sluggish kinetics of
photoelectrolysis of water and gas evolution reactions requires a large electrochemical
overpotential, which contributes to a low efficiency.
In this chapter, the fundamental basics of photoelectrochemical solar water
splitting along with its current challenges are reviewed. To overcome these challenges,
alternative approaches to more efficiently utilize sunlight such as photoelectrochemical
solar energy conversion and storage are introduced briefly. Among them, redox reaction-
assisted solar energy conversion and storage are considered one promising method to
achieve high efficiency, low-cost and non-intermittent solar energy utilization. At last, the
scope of this study and description of this dissertation are elaborated in details.
2.2 Operating Principles of the PEC
The principle of operating a typical PEC has been constantly reviewed in many
literatures[5, 8-12] and also explained in Figure 2-1. In general, photons of energy
exceeding that of the semiconductor bandgap (Eg) generate electron–hole pairs, which
are intrinsically separated by the electric field present in a very thin layer of the
semiconductor (called space charge region, SCR) at the interface. Under the influence of
the electrical field, the negative charge carriers move through the bulk of the n-type
7
semiconductor to the current collector and then to the external circuit. The positive holes
are driven to the surface where they are scavenged by a redox species (R) to its oxidized
form (O). The oxidized form O is reduced back to R on the counter electrode by the
electrons that re-enter the cell from the external circuit. This type of PEC is called
regenerative cell that converts light to electric power leaving no net chemical change
behind but heat. However, if the redox species is water, water will be cleaved by sunlight,
or photoelectrolysis, as water is oxidized by holes to oxygen at the semiconductor
photoelectrode and reduced by electrons to hydrogen at the other electrode. Because of
this promising merit, water-splitting by PEC is considered to be the most cost-effective
renewable approach for hydrogen production.[3, 5, 13]
Figure 2-1 Operation principle of photoelectrochemical cells based on n-type
semiconductor. a) Regenerative-type cell producing electric current from sunlight; b) A
cell that generates a chemical fuel, hydrogen, through the photo-cleavage of water.
2.3 Photoelectrochemical Water Splitting
The free energy change for the conversion of one molecule of H2O to H2 and 1/2
O2 under standard conditions is ΔG= 237.2 kJ/mol, which, according to the Nernst
8
equation, corresponds to ΔE°=1.23 V per electron transferred. To use a semiconductor
and drive this reaction with light, the semiconductor must absorb radiant light with photon
energy >1.23 eV (equal to wavelengths of ~1000 nm and shorter) and convert the energy
into H2 and O2. This process must generate two electron-hole pairs per molecule of
H2.[12, 13] Thermodynamically, a semiconductor must have a band gap energy (Eg)
large enough to split water, and a conduction band-edge energy (Ecb) and valence band-
edge energy (Evb) that straddle the electrochemical potentials of E° (H+/H2) and E°
(O2/H2O), to drive the hydrogen evolution reaction (HER) and oxygen evolution reaction
(OER) using the electrons/holes generated under illumination.
2.3.1 Thermodynamic Requirement
Figure 2-2 Band positions of common semiconductor materials in contact with aqueous
electrolyte at pH=1. Two red dashed lines indicate hydrogen and oxygen evolution
reaction potential. Several redox couple standard reduction potentials are also
represented against NHE level.
9
Figure 2-2 lists band positions of common semiconductor materials in contact
with aqueous electrolyte at pH=1 and the energy scale is indicated in eV using either
normal hydrogen electrode(NHE) or vacuum level as reference.[5] Two red dashed lines
represent the standard electrochemical evolution potential of H2 and O2 respectively. The
overall photocatalytic reaction of water splitting can only take place when the energy of
the absorbed photons is equal to or larger than the water splitting threshold energy of
1.23 eV which is calculated by equation 2.1 and 2.2, wherein ΔG0 and ΔE
0 are Gibbs free
energy change and standard electrochemical potential of the reaction under standard
conditions, n is the number of charge involved in the reaction, and F is the Faraday
constant. The mathematical representation below is referred to as the thermodynamic
requirement for H2 production by photoelectrolysis of water.
𝐻2𝑂 + 𝜐 → 𝐻2 +1
2𝑂2..........................................2.1
𝛥𝐺0 = −𝑛𝐹 ∙ 𝛥𝐸0..................................................2.2
In other words, the conduction band and valence band edges of the
semiconductor material must straddle over the energy levels of H2 and O2 reduction
potential simultaneously in order to split water. As seen in Figure 2-2, the conduction
band edge of many semiconductors are more positive rather than negative position
compared to the standard reduction potential of H2, thus unable to produce H2 through
photoelectrolysis of water.
2.3.2 Kinetic Requirement
The above discussion addresses the thermodynamic requirement for
photoelectrolysis of water by PEC. However, for meaningful water splitting efficiency,
more important limitations that lie on reaction kinetics and thus requires significant
overpotential as driving force must be overcome as oxygen evolution reaction at the
cathode involves activation of multistep reactions. First, the semiconductor material must
10
be stable, corrosion-free, and non-reactive within selected electrolyte. Second and the
most important, charge transfer after they are separated from the interface of
semiconductor/liquid junction must be fast enough to prevent charge carrier
recombination and energy loss be reduced due to overpotential.
2.4 Photoelectrochemical Solar Energy Storage
PECs can generate not only electrical but also electrochemical energy.
Conversion of a regenerative PEC to a photoelectrochemical storage solar cell (PESC)
can incorporate several increasingly sophisticated solar energy conversion and storage
configurations.
Figure 2-3 presents an example of a PESC combining in situ electrochemical
storage and solar-conversion capabilities, providing continuous output insensitive to daily
variations in illumination. A high solar-to-electricity conversion efficiency (11.3%) was
demonstrated using a PEC consisting of a Cd(Se,Te)/Sx and Sn/SnS storage system,
resulting in a solar cell with a continuous output.[14] Under illumination, as seen in Figure
2-3(a), the photocurrent drives an external load. Simultaneously, a portion of the
photocurrent is used to electrochemically reduce metal cations in the device storage half-
cell. Under dark or below a certain level of illumination, the storage compartment
spontaneously delivers power by metal oxidation, as seen in Figure 2-3(b).
11
Figure 2-3 Schematic of a photoelectrochemical solar cell combining both solar
conversion and storage capabilities. (a) Under illumination; (b) Under dark.
In light of such high solar-to-electrical (STE) conversion efficiency by realizing in-
situ photoelectrochemical processes, a series of photoelectrochemical solar energy
storage modes were either investigated or summarized by Stuart Licht.[15] Early studies
on PESC systems have been performed on a variety of two-electrode configurations and
important variations of these photoelectrochemical conversion and storage configurations
are summarized in Table 2-1 and schematically shown in Figure 2-4.
In each case, exposure to light drives separate redox couples and a current
through the external load. There is a net chemical change in the system, with an overall
increase in free energy. In the absence of illumination, the generated chemical change
drives a spontaneous discharge reaction. The electrochemical discharge induces a
reverse current. In each case in Table 2-1, exposure to light drives separate redox
couples and current through the external load. Consistent with Figure 2-1, in a
regenerative PEC, illumination drives work through an external load without inducing a
net change in the chemical composition of the system.
12
Table 2-1 List of exemplary photoelectrochemical conversion and storage modes using a
two-electrode configuration.
This compares with the two electrode PESC configurations shown in Figure 2-4.
Unlike a regenerative system, there is a net chemical change in the system, with an
overall increase in free energy. In the absence of illumination, the generated chemical
change drives a spontaneous discharge reaction. The electrochemical discharge induces
a reverse current. Utilizing two quasi-reversible chemical processes, changes taking
place in the system during illumination can be reversed in the dark. Similar to a
secondary battery, the system discharges producing an electric flow in the opposite
direction and the system gradually returns to the same original chemical state.
Figure 2-4 Cell structure of a PESC with (left) insoluble redox couple and (right) soluble
redox couple.
13
However, both cells shown in Figure 2-4 have some disadvantages. Firstly, the
redox species may chemically react with and impair the active materials of the
photoelectrode. Furthermore, during the discharge process, the photoelectrode is
kinetically unsuited to perform as an electrode for a reduction reaction in the absence of
illumination. For the photoelectrode to perform efficiently during illumination (charging),
such reduction process should be inhibited to minimize losses of back reaction which
essentially cancel the photo-oxidation.
Hence, the same photoelectrode cannot efficiently fulfill the dual role of being
kinetically sluggish to reduction upon illumination and yet being kinetically facile to the
same reduction during discharge under dark. Meanwhile, the configuration represented in
Figure 2-4 has another disadvantage, i.e., the disparity between the small surface area
needed to minimize the losses due to the reverse reactions and the large surface area
necessary to minimize storage polarization losses to maximize storage capacity.[16]
The above-mentioned disadvantages of using two-electrode configuration in
PESC may be overcome by considering a third electrode (symbol A) in the cell as shown
in Figure 2-5. In Figure 2-5, the switches E and F are generally alternated during charge
and discharge. During the charging, only switch E may be closed, facilitating the storage
process, and during discharge, E is kept open while F is closed. In this case, chemical
changes that take place during the storage process are reversed, and a current flow is
maintained from the storage electrode to a third (counter) electrode that is kept in the first
compartment. To minimize polarization losses during the discharge, this third electrode
should be kinetically active to the redox couple used in the first compartment.
14
Figure 2-5 Cell structure of a PESC with a third electrode (counter electrode) in the
photoelectrode compartment. P: Photoelectrode; A: Counter electrode; M: Membrane; S:
Storage electrode; EL: Electrolyte; F: Electrical Switch and L: Load.
Still an improved situation would have both switches closed all the time. In this
case, electric current flows from the photoelectrode to both the counter and storage
electrodes. The system is energetically tuned such that when illumination is available, a
significant fraction of the converted energy flows to the storage electrode. Under dark or
diminished illumination, the storage electrode begins to discharge, driving continuous
current through the load. In this system, a proper balance should be maintained between
the potential of the solar energy-conversion process and the electrochemical potential of
the storage process. There may be residual electric flow through the photoelectrode
during dark cell discharge, as the photoelectrode is sluggish, but not entirely passive, to a
15
reduction process. This can be corrected by inserting a diode between the
photoelectrode and the outer circuit.
2.4.1 Redox-assisted Solar Energy Conversion and Storage
Redox species, such as iodine, bromine, iron, cerium etc. in the electrolyte have
been reported[17-21] to improve photocatalytic reaction rates for H2 reduction/O2
reduction half-cell reactions significantly and thus water splitting efficiency. These redox
species serve as either electron donor or electron acceptor at the
semiconductor/electrolyte interface to quickly scavenge charge carriers and enhance
charge transfer due to their fast reaction kinetics.
Teruhisa Ohno and Michio Matsumura[20] discovered that photocatalytic
oxidation of water on TiO2 (rutile) powder proceeded with a fairly high efficiency (~9%)
when iron (III) ions were used as the electron acceptor, which is in marked contrast to
other reversible photocatalytic reactions, whose reaction rates decelerate as a result of
the back reactions of the products on the photocatalysts. The efficient oxidation of water
in the presence of iron (II) ions is attributed to preferential adsorption of iron (II) ions on
TiO2 over iron (III) ions, which enabled efficient oxidation of water.
The same group[17] also demonstrated a continuous water-splitting process into
hydrogen and oxygen under photoirridation using two redox couples on suspended TiO2
particles. In such a system, the reduction of water to hydrogen was realized by bromide
ions, which were oxidized to bromine. On the other hand, oxygen was created through
oxidation of water by Fe3+
ions, which were then reduced to Fe2+
ions.
16
Figure 2-6 Schematic of the photocatalytic generation of H2 under the proposed hole
shuttle mechanism.
More recently, the collaboration[22] between a group of scientists from Germany
and UK landed a remarkable increase in the H2 generation rate on chalcogenide
semiconductor nanocrystals by employing a OH-/OH˙ redox couple to efficiently relay the
hole from the semiconductor to the scavenger.
The proposed two-step mechanism shown in Figure 2-6 of hole transfer
employing a redox shuttle leads to a substantial improvement in the photocatalytic
generation of hydrogen by the metal-semiconductor nanocomposites, up to 53% of
external and 71% of internal quantum yield under 447 nm laser illumination, without the
use of noble metal co-catalysts and protection against photo-oxidation of the
photocatalyst. They concluded that the enhancement relies on much better mobility of the
small molecular shuttle OH-/OH˙, and its efficient reactions with the semiconductor and
the sacrificial agent, which implies that replacing the rate-limiting step with two faster
17
processes involving a redox mediator could be a viable approach to efficient
photocatalysis.
2.5 Scope of Study and Objectives
In this study, we propose to use vanadium redox species (in the form of
VO2+
/VO+ and V
3+/V
2+), the fundamental components of a vanadium redox battery(VRB)
to convert and store solar energy due to the following unique characteristics:
i. Vanadium redox species and its behavior have been well studied back in
1980’s and results lead to successful commercial products.
ii. Vanadium redox species have much faster kinetics in electrochemical
reaction compared to H+/H2 and O2/H2O redox pairs in photocatalytic
water splitting and therefore are prone to yield higher voltage and current
efficiencies.
Our previous study demonstrated high reaction kinetics for VO2+
/VO2+
redox
couples shown in Figure 2-7. The vanadium redox couple shows much higher current at
the same overpotential. The exchange current density (i0, marked by the arrows in Figure
2-7 lower graph) is calculated to be about six orders of magnitude higher than that of
oxygen reduction reaction. Such a large i0 would significantly compensate the
overpotential (η) and thus improve solar energy conversion efficiency.
18
Figure 2-7 Kinetics comparison of reduction reaction of O2/H2O and VO2+
/VO+ in 3 M
H2SO4.
iii. Simultaneous in-situ solar energy conversion and storage could be
achieved. Figure 2-8 demonstrates such energy conversion approach, in
a so-called all-vanadium photoelectrochemical cell as we developed in
this dissertation.
19
Figure 2-8 Schematic of the solar energy storage approach via photocatalytic redox
reactions.
In such a PEC, a semiconductor material acts as photocatalyst/photoanode to
convert and store solar energy into chemical energy by promoting oxidation/reduction
reactions with two redox species, e.g., VO2+
/VO2+
and V3+
/V2+
in the electrolyte. During
charge, the holes and electrons will oxidize and reduce different redox components at the
anode and cathode, respectively, and thus achieve the goal of energy conversion. During
discharge under dark, the energy stored in the redox species will be released by
reversible electrochemical reactions if a load is connected.
Compared to conventional water splitting by a PEC, the proposed all-vanadium
PEC excels in that the charge and discharge process can occur either separately or
simultaneously, thus eliminating issues related to intermittent nature of solar illumination
as a function of time or weather. Given the high reaction kinetics of vanadium redox
species, a great enhancement on solar energy conversion efficiency would be expected if
successful.
20
2.6 Description of This Dissertation
This dissertation is organized into the following chapters:
In Chapter 1, three major approaches to utilize solar energy are briefly
introduced. The advantages and disadvantages of PECs, the most promising approach to
generate either electricity or chemical fuel such as hydrogen, were elaborated and
followed by the general description of the work demonstrated in this dissertation to
conquer the major issues associated with conventional PECs.
In Chapter 2, fundamental aspects of PECs, including operating principles,
thermodynamic and kinetic requirements in terms of photoelectrochemical water splitting
will be presented. In addition, early work in the field with respect to photoelectrochemical
solar energy storage will be summarized and the concept of redox-assisted solar energy
conversion and storage will be disclosed. Finally, the scope and objectives of this study
are proposed in details based upon our preliminary results.
In Chapter 3, photoelectrochemical studies have been performed on TiO2
electrode and WO3/TiO2 hybrid electrode in an all-vanadium PEC to prove the concept of
the proposed research. During the study, a photocharging/discharging process enabled
by WO3 acting as an effective electron storage reservoir was discovered.
In Chapter 4, the interaction between WO3/TiO2 hybrid electrode and vanadium
redox reactions will be deeply investigated in an all-vanadium PEC using multiple
electrochemical methods. A model aiming to explain the observed experimental results is
proposed.
In Chapter 5, optimization of the photoelectrochemical performance of the all-
vanadium PEC using a novel supporting electrolyte, methanesulfonic acid, has been
performed in comparison with the system using sulfuric acid. The quantification that
21
defines the cell performance such as Faradaic efficiency and IPCE of the cell has been
conducted.
In Chapter 6, all experimental results achieved from Chapter 3 to Chapter 5 in
this dissertation are concluded.
22
Chapter 3
Primary Study of All-Vanadium PEC Using Different Photoelectrodes
3.1 Introduction
In this chapter, common photocatalysts, e.g., TiO2 and WO3, were employed to
benchmark the photocatalytic properties of the proposed all-vanadium PEC. Not only we
conducted the proof-of-the-concept experiments using a typical TiO2 photocatalyst to test
the system, but also an electron storage mechanism was discovered on WO3/TiO2 hybrid
electrode preliminarily.
3.2 Experiments
3.2.1 Electrode and Electrolyte Preparation
Three types of photoelectrodes, TiO2, WO3, and WO3/TiO2 hybrid (12 wt% WO3),
were fabricated and used throughout the experiment. To fabricate the TiO2 and hybrid
electrodes, 0.997 g Degussa P25 TiO2 (VP AEROPERL® by Evonik), 0.497 g ethyl
cellulose (48.0-49.5%, Sigma-Aldrich USA), 0.124 g polyvindylidene fluoride (PVDF)
powder (Kynar Flex 2801-00 by Arkema Group), 2.501 g α-terpineol (laboratory grade,
Fisher Scientific USA) were mixed under constant stirring at 80°C for 1 h to obtain a
uniform TiO2 or hybrid slurry with the addition of 0.374 g tungstic acid (Alfa Aesar, USA)
and 0.993 g hydrogen peroxide (35 %, Alfa Aesar USA). Then the slurry was deposited
on a pre-cut square fluorine doped tin oxide (FTO) (sheet resistance 6-8 Ω/□, Pilkington
USA) using a doctor blade. The FTO substrate was pre-washed with acetone (99.7%,
Fisher Scientific USA), methanol (99.8%, Fisher Scientific USA), and deionized (DI)
water, before being blow-dried and then further dried in an oven at 120°C for 1h. After the
deposition, the obtained coating was dried in the oven for 1 h and then calcined under air
flow at 325°C for 5 min, 375°C for 5 min, 450°C for 15 min, and finally 500°C for 15 min.
23
The fabrication of WO3 electrode was performed in a similar manner. After the
fabrication, the active electrode area was approximately 6.45 cm2.
Four types of electrolytes, 3 M H2SO4, 0.1 M vanadium(IV, VO2+
) in 3 M H2SO4,
0.01 M vanadium(IV) in 3 M H2SO4,and 0.01 M vanadium(III, V3+
) in 3 M H2SO4, were
used in the experiments. The first three electrolytes were prepared by dissolving
stoichiometric amount of H2SO4 (96.6%, J.T. Baker USA) in DI water with or without
vanadium(IV) sulfate oxide hydrate (VOSO4·xH2O) (99.9%, Alfa Aesar USA). The number
of hydrate in VOSO4·xH2O was pre-determined by thermogravimetric analysis. The
prepared 0.01 M vanadium(IV)-sulfuric acid and 0.1 M vanadium(IV)-sulfuric acid
electrolytes appear light blue and blue respectively. The 0.01 M vanadium(III) electrolyte
was obtained by electrochemically reducing the prepared vanadium(IV)-H2SO4 solution in
an electrochemical cell at a constant current density of 3 mA/cm2 using a potentiostat
(Princeton Applied Research, PARSTAT 2273) until the potential reached 1.6 V. During
the reduction, the electrolyte was protected by N2 throughout the experiment to prevent
oxidation of vanadium(III) species. The obtained 0.01 M vanadium (III)-sulfuric acid
solution appeared light green.
3.2.2 Cell Design and Assembly
Linear sweep voltammetry (LSV), cyclic voltammetry (CV), and zero-resistance
ammetry (ZRA) were conducted in either a half-cell configuration(one chamber) or full-
cell configuration (two chambers filled with VO2+
and V3+
electrolytes, separated by a
Nafion117 membrane) with their schematics shown in Figure 3-1. In both cell
configurations, the photoelectrode serves as the working electrode (WE) while a platinum
mesh and Ag/AgCl electrode serve as the counter electrode (CE) and reference
electrodes (RE), respectively. The voltage scan ranged from -0.197 to 1.3 V with 0.1 V
interval of dark/illumination. The overall duration for ZRA measurement was180 s with 20
24
s intervals of dark/illumination. The solar irradiation was created using an ozone-free
solar simulator system (Newport USA) coupled with an AM 1.5 global filter (Newport
USA) and calibrated using a standard Si photodiode (Newport USA).
Figure 3-1 The representative schematics of (a) half-cell configuration and (b) full-cell
configuration.
3.2.3 Material Characterization
The crystallographic information of the electrodes was determined by XRD
(Siemens, 810-M340-32-C3000) at a scan rate 0.01° s-1
between 20°-80° with a dwell
time of 1 s. Scanning electron microscopy (Hitachi S-3000N variable pressure SEM) was
used to examine the microstructure of photoelectrodes. The energy dispersive
spectroscopy (EDS) mapping was also performed to reveal elemental composition of the
sample. The UV-Vis diffuse reflectance spectra of the samples were obtained on a
JASCO V-570 spectrophotometer while Raman spectra were collected on a PerkinElmer
DXR Raman microscope.
25
Figure 3-2 XRD spectra of the TiO2 and WO3/TiO2 (12 wt% WO3) hybrid electrodes.
Figure 3-2 presents the XRD spectra of the TiO2 and WO3/TiO2 (12 wt% WO3)
hybrid electrodes. The JCPDS standards for anatase TiO2 (#21-1272), rutile TiO2 (#76-
1940), and monoclinic WO3 (#83-0950) were also shown as reference. It is found that
only anatase and rutile phases of TiO2 existed in both samples with the former being
predominant. This finding is in agreement with the revealed phase information of the
commercial Degussa P25 product. Also, only monoclinic phase of WO3 was found in the
hybrid sample, and no noticeable impurity phase such as bromite TiO2 or orthorhombic or
tetragonal WO3 existed by cross-referencing the XRD spectra with the standards.
26
Figure 3-3 Raman spectra of the TiO2 and WO3/TiO2 (12 wt% WO3) hybrid electrodes.
Raman spectroscopy was employed to verify the phases of TiO2 and WO3 along
with carbon species in the photoelectrode. In Figure 3-3, five Raman active modes near
146, 197, 397, 515, and 633 cm-1
are clearly observed for both samples. These peaks
are characteristic vibration modes of a TiO2 anatase phase. No TiO2 rutile phase is
observed even though it was reported to appear near 455 cm-1
.[23] The vibration modes
of graphitic carbon (1200-1600 cm-1
) were not observed (results not shown), which
indicates that carbon species were all removed during the sintering process. The band
near 270 cm-1
is emerging from the O-W-O deformation vibration and bands around 713
and 806 cm-1
are attributed to the O-W-O stretching vibrations.[23, 24] All these bands
belong to the characteristic vibration modes of monoclinic WO3 and all Raman results are
consistent with XRD patterns revealed in Figure 3-2.
27
Figure 3-4 SEM images of sintered TiO2 sample in (A) lower and (B) higher magnification,
together with (C) unsintered TiO2 powder.
Figure 3-4 demonstrates morphologies of the sintered TiO2 photoelectrode in
comparison with the unsintered TiO2 powder. It is seen from Figure 3-4A that a
continuous porous network of TiO2 is formed by randomly-organized spherical particles
ranging from 3-25 µm and no explicit crack was found in the sample. A close view of the
particles shown in Figure 3-4B reveals that the sintered TiO2 sample has its particles
adjoined together due to heat treatment and craters/irregularities of various sizes are
found to exist on the surface rendering it very rough. On the contrary, the unsintered TiO2
powders (Figure 3-4C) were composed of many isolated spheres with relatively smooth
surface although particle size remains unchanged.
28
Figure 3-5 SEM image of the WO3/TiO2 (12 wt% WO3) hybrid electrodes. A higher-
magnification image is shown in the inset.
SEM images were also taken on the hybrid electrode in Figure 3-5 in order to
compare with the pure TiO2 electrode. Possessing a similar porous network to the pure
TiO2 electrode as demonstrated in Figure 3-4A, the hybrid sample, however, exhibites a
binomial distribution of particles consisting mainly of large TiO2 particles along with small
WO3 particles of 1-2 μm that are embedded discretely in the matrix. TiO2 particles are
adjoined with not only themselves but also their neighboring WO3 particles. Many mini-
cracks are found to exist in the matrix. As seen in the inset image, WO3 particles tend to
agglomerate more easily than TiO2 due to their much smaller dimensions and they were
also found to be scattered on the surface of TiO2 particles. Note that the size of TiO2
particles in the hybrid sample remains unchanged compared to that of the pure TiO2
sample.
29
Figure 3-6 EDS mapping result of the WO3/TiO2 (12 wt% WO3) hybrid electrode.
WO3 particles were found to distribute all over the matrix in the hybrid electrode,
according to EDS mapping result shown in Figure 3-6. Only three elements, Ti, W, and O
were found to exist in the sample analyzed by EDS and TiO2 was demonstrated to
dominate in the electrode over WO3 as indicated by rather distinguished Ti element on
the mapping graph. On the other hand, WO3 particles were also found to scatter around
and over the TiO2 particles and such intimate contact between the two semiconductors is
expected to facilitate charge carrier transfer between WO3 and TiO2, and reduce their
recombination after they are generated.
30
Figure 3-7 UV-Vis diffuse reflectance spectra of the TiO2 and WO3/TiO2 (12 wt% WO3)
hybrid electrodes.
The optical property of both the TiO2 and hybrid electrodes was studied by UV-
Vis diffuse reflectance spectroscopy and the spectra are revealed in Figure 3-7. For TiO2,
the absorption edge resides in near 380 nm as TiO2 is well known to absorb only the UV
light. The hybrid sample shows a clear red shift on the spectrum, extending the
absorption edge into the visible light region near 450 nm. Compared to the pure TiO2
sample, the hybrid one also demonstrates an improved absorption tail ranging from 500
to 800 nm, which is ascribed to the smaller band energy of 2.6-2.8 eV of WO3.[24-26]
31
3.3 Photoelectrochemical Study
3.3.1 Proof-of-the-concept Experiment
-0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
-0.2
0.0
0.2
0.4
0.6
0.8I,
mA
E vs Ag/AgCl, V
Dark, 3M H2SO
4
AM1.5, 3M H2SO
4
Dark, 0.1M VO2+
in 3M H2SO
4
AM1.5, 0.1M VO2+
in 3M H2SO
4
Figure 3-8 The linear sweep voltammograms of the TiO2 electrode under various
electrochemical conditions in a half-cell configuration. The scan rate was 5 mV/s.
The proof-of-the-concept experiment was conducted using the TiO2 electrode by
LSV under various electrochemical conditions in a half-cell configuration. The results
shown in Figure 3-8 clearly indicate that vanadium redox species (VO2+
in this
experiment) is capable of improving photocatalytic property of PEC significantly, in other
words, the feasibility of our hypothesis.
As seen in Figure 3-8, little difference in photoelectrochemical response of the
TiO2 photoelectrode is observed regardless of the presence of illumination when only
acid was employed as the electrolyte. Furthermore, the barely noticeable photocurrent
throughout the potential window under illumination manifests the well-known wide
32
bandgap of TiO2[7] and its notorious electron-hole pair recombination at the
semiconductor/liquid junction interface. However, the addition of VO2+
species in the
electrolyte abruptly changes the voltammetry response as depicted in Figure 3-8,
especially under illumination. Two distinct photoelectrochemical behaviors are seen when
VO2+
species participate the electrochemical and/or photoelectrochemical reactions. At
potentials less positive than 1.0 V, TiO2 shows a typical n-type semiconductor
photoelectrochemical behavior in 0.1 M-3 M H2SO4 electrolyte under AM 1.5 illumination,
except that the photocurrent achieved is magnificently enhanced on average 14 times
higher than that with only 3 M H2SO4 as the electrolyte from 0.0 to 1.0 V. The significant
improvement in photoresponse is believed to be induced by a synergistic effect of
photocatalysis of TiO2 and oxidation of VO2+
species. Fast electrochemical kinetics of
VO2+
oxidation (at least six orders of magnitude higher than the oxygen evolution) is
believed to quickly scavenge holes. This thus minimizes charge recombination and
improves photocatalytic electron-transfer efficiency at the semiconductor interface.[9] The
other interesting electrochemical and/or photoelectrochemical behavior of TiO2 in 0.1 M
V-3 M H2SO4 electrolyte appears at potentials more positive than 1.0 V. The current
increases abruptly as the potential increases regardless of the illumination, which may be
ascribed to tunneling effect of electrons across the semiconductor/liquid junction in both
dark and illumination conditions.
In classic photoelectrochemistry of semiconductors,[5, 6, 8-10, 27-29] when light
illuminates upon a n-type semiconductor immersed in a liquid solution containing a redox
species, oxidative electron transfer reaction occurs at the semiconductor interface once
the electrolyte redox potential is positioned below that of the photogenerated hole at edge
of the bended valence band of the semiconductor. As a result, the redox species will be
oxidized photo-electrochemically. In our study, as the potential scale for the upper edge
33
of valence band of TiO2 against NHE (Normal Hydrogen Electrode) in contact with
aqueous solution at pH=1 is approximately 2.8 V[5] and the standard potential of V4+
/V5+
redox is 1 V,[30] it is obviously facile for the VO2+
species to be quickly oxidized by
scavenging the holes at the semiconductor surface according to the principle mentioned
above, which consequently reduces the electron-hole pair recombination and enhances
the photoresponse. Note that the instability of voltammetry signals between 0.2 to 0.8 V
is believed to be caused by the possible competing scavenging process of
photogenerated holes among a series of vanadium species complex[31-33] though V(IV)
species exists predominantly in the form of VO2+
in pH<1 acidic solution.[34]
The effect of vanadium species concentration on the current under both dark
and illumination conditions was also studied and the results are shown in Figure 3-9. Two
photocurrent ratios at both 0.5 and 1.0 V, including those between oxidation of VO2+
(V)
and water (W) (Figure 3-9A) and between those under illumination and dark (Figure 3-
9B) for vanadium species, are shown. Figure 3-9A clearly shows that at a lower potential
(0.5 V) the photocurrent ratio between oxidation of VO2+
(V) and water (W)slightly
increases with increasing VO2+
concentration (below 0.5 M). However, the ratio becomes
saturated and even reduces when the VO2+
concentration is beyond 0.5 M, which is
suspected to be caused by incorporation of vanadium species into TiO2 lattice and
therefore deactivation of the semiconductor. At a higher potential (1.0 V vs. Ag/AgCl), the
ratio improves significantly suggesting fast photoelectrochemical kinetics of the VO2+
species. On the other hand, the photocurrent ratio between those under AM 1.5 and dark
shown in Figure 3-9B decreases with VO2+
concentration, suggesting that at higher
concentrations the oxidation current from purely electrochemical reaction may dominate
over the photoelectrochemical process. The above-mentioned results indicate that the
proposed vanadium ions are very photo/electrochemically active. To effectively reveal the
34
photoelectrochemical storage/conversion using these redox species, lower concentration
and lower voltage (particularly, no external bias) are preferred.
0.0 0.2 0.4 0.6 0.8 1.0
0
20
40
60
80
100
IvA
M1.5
/Iw
AM
1.5
@ 1.0 V
(A)
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
(A)
IvA
M1.5
/Iw
AM
1.5
VO2+
Concentration, M
@ 0.5 V
0.0 0.2 0.4 0.6 0.8 1.0
0
5
10
15
20
25
IvA
M1
.5/IvD
ark
@ 1.0 V
(B)
0.0 0.2 0.4 0.6 0.8 1.0
0
5
10
15
20
25
(B)
IvA
M1
.5/IvD
ark
VO2+
Concentration, M
@ 0.5 V
Figure 3-9 The plot of (a) IvAM1.5/I
wAM1.5 and (b) I
vAM1.5/I
vDark as a function of vanadium
species concentration at 0.5 V and 1V.
3.3.2 Photoelectrochemical Performance of the Hybrid Electrode
The above results have clearly demonstrated significant ability of vanadium
redox species in enhancing photoelectrochemical performance of the PEC and thereof
the feasibility of the proposed hypothesis. To further improve the photocatalytic property
of the PEC, we also fabricated WO3/TiO2 hybrid electrode (12 wt% WO3) and compared
its photocatalytic performance with TiO2 electrode under various electrochemical
circumstances.
35
Figure 3-10 Linear sweep voltammograms of three different electrodes in 3 M H2SO4
electrolyte in a half-cell configuration under alternate dark and illumination conditions.
The scan rate was 5 mV/s.
To compare the photoelectrochemical performance of different electrodes, LSV
was firstly employed in a half-cell configuration in the experiment and the results are
shown in Figure 3-10. The results show insignificant photocurrents of TiO2 and WO3
electrodes under AM 1.5 illumination. However, the hybrid electrode exhibited much more
noticeable photoelectrochemical response corresponding to a higher reaction rate of the
following equation particularly at
higher potentials.
𝑉𝑂2+ +𝐻2𝑂 → 𝑉𝑂+ + 𝑒− + 2𝐻+ (E0=1.0 V)....................3.1
Besides, the photocurrent of TiO2 and WO3 electrodes remained almost
unchanged at higher potentials under illumination whereas the photocurrent of the hybrid
36
electrode was enhanced at least by a factor of 3-4. The enhanced photoresponse is
clearly indicative of mitigated charge recombination and thus improved
photoelectrochemical reaction rate. It was well reported in literatures[24-26, 35] that
monoclinic WO3 has a more positive valence band (VB) than TiO2 and photogenerated
holes tend to migrate from its VB to that of TiO2 and then to the catalyst surface under
illumination. Electrons, however, favor migrating from the conduction band (CB) of TiO2
to that of WO3 and then to the bulk of the semiconductor due to the more negative
potential of TiO2 CB. As a result, the photogenerated charge carriers would tend to
separate rather than recombine in the bulk, leading to a significant improvement in
photocurrent. What's also worth mentioning in Figure 3-10 is that a broad peak emerges
around 0.1 V on the anodic scan. This is believed due to
electrochromism/photochromism of WO3 under certain electrochemical conditions and
the detailed study of this observation is given later.
As evidenced in our previous work,[36] fast redox reactions using vanadium
redox couples are capable of boosting photoelectrochemical performance via quickly
scavenging holes at the semiconductor/electrolyte interface and thus mitigating
recombination of charge carriers after they are generated. However, our previous study
of such effect relied on cyclic voltammetry that requires an external bias applied to the
cell. In the study performed herein, zero-resistance ammetry (ZRA) method was adopted
in the experiments to investigate the photoelectrochemical properties of different
electrodes and the results are given in Figure 3-11.
37
Figure 3-11 Photocurrents of different photoelectrodes using ZRA method in (A) 3 M
H2SO4 electrolyte and (B) various vanadium electrolytes in different cell configurations.
The labels in Figure 3-11B designate: 1: TiO2 in 0.01 M V(IV)-H2SO4, half-cell; 2: TiO2 in
0.01 M V-H2SO4, full-cell; 3: Hybrid in 0.01 M V(IV)-H2SO4, half-cell; and 4: Hybrid in 0.01
M V-H2SO4, full-cell.
Figure 3-11A compares the photoresponse of three different photoelectrodes in 3
M H2SO4 electrolyte with a half-cell configuration while Figure 3-11B compares TiO2 and
38
hybrid electrodes with a full-cell configuration. As revealed in both graphs, the hybrid
electrode shows a more pronounced photocurrent than TiO2 or WO3 electrode regardless
of the electrolyte and cell configuration. Such improvement is ascribed to facilitated
charge carrier transfer from WO3 to TiO2 due to their band structure interplay, as
discussed earlier.
On the other hand, the vanadium redox pairs have a great contribution toward
photocurrent enhancement regardless of the electrode used. This enhancement is even
more conspicuous when two vanadium redox species, i.e., VO2+
and V3+
, and a full-cell
configuration were coupled with each other. As indicated by the black and red curves in
Figure 3-11A, the photoelectrochemical performance of both TiO2 and hybrid electrodes
in 3 M H2SO4 was unsatisfactory in terms of photocurrent magnitude. However, the
addition of vanadium redox species even with a small molarity of 0.01 M (black and green
curves in Figure 3-11B) at least doubled the photocurrent for both electrodes, which
again proves the fast reaction kinetics of vanadium redox pairs.
Furthermore, when two pairs of vanadium redox species were used in separate
chambers with a full-cell configuration, the photocurrent was significantly boosted for both
electrodes especially the hybrid one, as clearly demonstrated by the blue and red curves
in Figure 3-11B. Instead of the hydrogen evolution reaction taking place with a half-cell
configuration, reduction reaction of the following equation occurred at the counter
electrode in a full-cell configuration, which is responsible for the improved photocurrent.
𝑉3+ + 𝑒− → 𝑉2+ (E0=-0.26 V)....................3.2
3.3.3 Preliminary Electron Storage Study by Hybrid Electrode
It is also interesting to note in Figure 3-11 that the hybrid electrode showed a
noticeable dark current after the illumination was switched off, which indicates its greater
ability to store electrons and release of the stored electrons. We ascribe this occurrence
39
to the formation of tungsten bronze induced by the photocharge/discharge process under
higher energy irradiation on the WO3 surface. It is known that WO3 tends to be
electrochemically reduced to tungsten bronze by combining electrons and protons (H+)
under UV irradiation as depicted in the following equation: [37-39]
Due to such reaction, the surface of the photocatalysts, i.e., TiO2 and WO3, will
be covered with HxWO3, which prevents the photogenerated holes from moving from the
valence band of WO3 to that of TiO2 and further to semiconductor/liquid interface to
oxidize VO2+
ions. This thus compromises photocatalysis of the semiconductor materials.
When irradiation was turned off, tungsten bronze undergoes discharge/self-discharge,
which transforms itself back to WO3. As a result, more semiconductor material will rejoin
the photocatalysis process. Furthermore, the transfer of the photogenerated holes may
be again possible from the WO3 valence band to the TiO2 valence band, thus enhancing
the photocurrent. This photocharge/discharge process was confirmed by the blue
appearance of the electrode under AM 1.5 irradiation and disappearance under dark
through certain period of time. However, different from the electrode’s behavior in Figure
3-11A, charge carriers can be quickly consumed by the vanadium redox species in the
electrolyte especially in a full-cell configuration and thus renders a negligible dark current.
3.3......................WOHxHxeWO 3x3
discharge
ephotocharg
40
Figure 3-12 CV of the WO3 electrode in 3 M H2SO4 electrolyte under dark and AM 1.5
illumination in a half-cell configuration. The scan rate was 20 mV/s.
It has been reported[26, 40-42] that storage of photogenerated electrons in WO3
is achievable under one or more of the following conditions: (i) the presence of UV
illumination and a hole scavenger, (ii) application of a negative bias, and/or (iii) use of an
external source of electrons. To further study the formation of tungsten bronze, multiple
CV scans were conducted on the pure WO3 electrode in 3 M H2SO4 electrolyte under
alternate dark/illumination to eliminate the effect of other factors, such as hole
scavengers and external electrons provided by TiO2 for instance. The first nine cycles
were conducted under dark and all CV curves overlapped with each other as shown in
Figure 3-12. This result indicates that no detectable electrochemical reactions occurred
on the electrode in an acidic environment under dark conditions. At the 10th scan, when
AM 1.5 illumination was applied, the WO3 electrode showed a broad peak centered
41
around 0.1 V due to the formation of tungsten bronze from WO3 reacting with H+ ions and
the photogenerated electrons.[36, 40, 42] The anodic current induced by the illumination
is appreciable throughout the potential window and the cathodic current at potentials
below 0.05 V vs. Ag/AgCl results from the decomposition of tungsten bronze that
releases both H+ ions and electrons. These peaks disappeared when AM 1.5 illumination
was turned off at the 11th scan and the current went back to the original value eventually
at the 15th scan, indicating a reversible behavior of the reactions.
The integration of a WO3/TiO2 hybrid electrode and vanadium redox pairs offers
a new approach to efficiently store solar energy. The vanadium redox pairs serve two
purposes: (i) quickly scavenge the charge carriers, mitigate their recombination and thus
boost the photocurrent, and (ii) shift the dynamic equilibrium between charge carrier
recombination and redox reaction toward the latter due to fast kinetics of these vanadium
redox couples. The hybrid electrode is able to not only effectively facilitate charge
transfer between the two semiconductors and enhance the light absorption, but also
serves as electron storage reservoir. The latter is critical to yield constant PEC reactions
that are responsible for the regenerative solar energy storage. The built-in electron
reservoir in the photoelectrode would result in stable electron storage and supply albeit
the intermittent nature of sunlight.
3.4 Conclusion
Three types of photoelectrodes, TiO2, WO3, and WO3/TiO2 hybrid electrode (12
wt% WO3) were fabricated and their photoresponse was investigated by CV and ZRA
under various electrochemical and photoelectrochemical conditions.
The proof-of-the-concept experiment was conducted using LSV in 0.1 M V-
H2SO4 electrolyte under AM 1.5 illumination and the results clearly reveal a significant
photoresponse enhancement compared to photolysis of water even with a pristine TiO2
42
electrode. This is attributed to the hole scavenging process initiated by quick redox
reaction and depressed charge recombination at semiconductor/liquid interface.
In order to further improve photoelectrochemical performance of the PEC, a
WO3/TiO2 hybrid electrode (12 wt% WO3) was fabricated and probed using CV and ZRA.
The hybrid electrode exhibited higher photoresponse than TiO2 electrode and this
improvement is ascribed to facilitated charge carrier transfer from WO3 to TiO2 due to
their band structure interplay.
Meanwhile, an all-vanadium photoelectrochemical storage cell has been
demonstrated by coupling two vanadium redox pairs instead of one with either TiO2 or
WO3/TiO2 hybrid electrode. The photoelectrochemical study performed shows the
photocurrent has been significantly enhanced with the assistance of two vanadium redox
pairs especially using a TiO2/WO3 hybrid electrode. It is believed that the vanadium redox
pairs play a key role in the regenerative solar energy storage due to their fast reaction
kinetics.
Various material characterization was also conducted on all used electrodes and
the results imply that an intimate contact between WO3 and TiO2 is key to enhancement
in charge carrier transfer, depression of recombination, and improvement in the
photocurrent.
Most important, a photocharge/discharge process was discovered on the WO3
under different energy level of irradiation, leading to a deteriorated photocurrent. This
interesting phenomenon is found to be associated with the formation/decomposition of
tungsten bronze under higher or lower energy of irradiation respectively. The preliminary
study of hybrid electrode in combination with vanadium redox reactions demonstrates
that WO3 therein functions as an effective electron storage reservoir, which would
43
potentially enable a unique dual solar/electron storage system albeit the intermittent
nature of sunlight.
44
Chapter 4
Electron Storage Study of Hybrid Electrode
4.1. Introduction
In chapter 3, we proved the great ability of vanadium redox species enhancing
the photoresponse of a PEC even with a pristine TiO2 photoelectrode. Furthermore, the
photoelectrochemical study with a hybrid electrode, i.e., WO3/TiO2, showed the unique
electrochromism/photochromism property of WO3 as an electron storage reservior to
store solar energy in certain conditions. Such energy storage ability is realized through
the formation of tungsten bronze by WO3 reacting with electrons and protons in an
electrochemical system under solar illumination.
In light of that, we combined the energy storage ability of WO3 with our newly
developed all-vanadium photoelectrochemical storage cell to explore the possibility of
photoelectrochemical solar energy conversion and storage, especially in the case of
intermittent sunlight in this chapter.
The combined system incorporates the conventional PEC, a vanadium redox-
flow battery (VRB), and a WO3/TiO2 hybrid photocatalyst. Zero-resistance ammetry (ZRA)
and electrochemical impedance spectroscopy (EIS) were employed to further study
photoelectrochemical response and reaction mechamism of this system in conversion
and storage of solar energy under both illumination and dark. The results indicate an
important synergy between electron storage and the all-vanadium electrolytes, which
potentially offers great reversibility, high-capacity electron storage, and significant
improvement in photocurrent.
4.2 Experiments
In a typical electrode fabrication procedure, TiO2 and WO3/TiO2 hybrid (various
WO3 loadings) photoelectrodes with active area of 6.45 cm2 were fabricated and used
45
throughout the experiment. To fabricate TiO2 and hybrid electrodes, 1.00 g Degussa P25
TiO2 (VP AEROPERL® by Evonik), 2.50 g α-terpineol (laboratory grade, Fisher Scientific
USA) were mixed under constant stirring at 80°C for 1 hr to obtain a uniform TiO2 or
hybrid slurry with the addition of tungstic acid (Alfa Aesar USA). Then the slurry was
deposited on a pre-cut square-shaped fluorine doped tin oxide (FTO) (Pilkington USA)
using a doctor blade. The FTO substrate was pre-washed with acetone (99.7%, Fisher
Scientific USA), methanol (99.8%, Fisher Scientific USA), and deionized (DI) water,
before being blow-dried and then further dried in an oven at 120°C for 1 hr. The obtained
coating was subsequently calcined with air flow at 500°C for 90 min.
Three types of electrolytes, 3 M H2SO4, 0.01 M vanadium(IV, VO2+
) in 3 M
H2SO4, and 0.01 M vanadium(III, V3+
) in 3 M H2SO4, were used in the experiments. The
first two electrolytes were prepared by dissolving H2SO4 (96.6%, J.T. Baker USA) in DI
water with or without vanadium(IV) sulfate oxide hydrate (VOSO4·xH2O) (99.9%, Alfa
Aesar USA). The number of water in VOSO4·xH2O was determined by thermogravimetric
analysis. The prepared vanadium(IV)-sulfuric acid solution appeared light blue. The 0.01
M vanadium(III) electrolyte was obtained by electrochemically reducing the prepared
vanadium(IV)-H2SO4 solution in an electrolytic cell at constant current density of 3
mA/cm2 using a potentiostat (Princeton Applied Research, PARSTAT 2273) until the
potential reached 1.6 V. During the reduction, the electrolyte was protected by N2 to
prevent oxidation of the vanadium(III) species. The obtained vanadium(III)-sulfuric acid
solution appeared light green.
46
20 30 40 50 60 70 80 2
anatase TiO2
rutile TiO2
mono WO3
a
b
c
d
Figure 4-1 XRD spectra of (a) TiO2 electrode, and hybrid electrodes with various WO3
loadings: (b) 6 wt%, (c) 12 wt%, and (d) 24 wt%.
The crystallographic information of all electrodes was determined by XRD
(Siemens, 810-M340-32-C3000) at a scan rate 0.01°s
-1 between 20
°-80
° with a dwell time
of 1 s. No impurity other than anatase phase and rutile phase TiO2, and monoclinic WO3
was discovered from the XRD spectra (Figure 4-1).
47
Figure 4-2 SEM images of (A) TiO2 electrode and hybrid electrodes with different WO3
loadings (B) 6 wt%, (C) 12 wt%, and (D) 24 wt%. High-magnification images are shown in
the inset for each figure. The scale bars in low-magnification and high-magnification
images represent 300 um and 10 um respectively.
Scanning electron microscopy (Hitachi S-3000 N variable pressure SEM) was
used to examine the microstructure of all electrodes. The obtained SEM images (Figure
4-2) clearly show small WO3 clusters on TiO2 particles and mud cracks at higher WO3
loadings, which was in agreement with our previous findings.[36, 43]
Raman spectra were collected using a PerkinElmer DXR Raman microscope
from 100 to 2000 cm-1
. In the Raman spectra (Figure 4-3), no other peaks other than
anatase TiO2 and monoclinic WO3 were observed. Five Raman active modes near 146,
197, 397, 515, and 633 cm-1
are assigned to characteristic vibration of anatase TiO2.[43]
48
The peaks near 270, 326, 713 and 806 cm-1
belong to characteristic vibration modes of
monoclinic WO3.
200 400 600 800 1000
Raman shift (cm-1)
A
AAA
Before ZRAA
A: Anatase TiO2
MW: Monoclinic WO3
MW
MW
MW
After ZRA
MW
Figure 4-3 Raman Spectra of the hybrid electrode (24 wt% WO3) before and after ZRA
experiments in 3 M H2SO4 electrolyte.
The photoelectrochemical property of all electrodes was studied by CV, ZRA and
EIS in a two-chamber, three-electrode electrochemical cell, where the photoelectrode
served as the working electrode (WE), a platinum mesh and Ag/AgCl electrode served as
the counter electrode (CE) and reference electrodes (RE), respectively. Details of the
experimental setup have been described in our previous work.[43, 44] Tests using zero-
resistance ammetry (ZRA) were conducted up to almost 4 hrs with either 3 M H2SO4 or
0.01 M vanadium in 3 M H2SO4 electrolyte under dark and AM 1.5 conditions. In a typical
experiment, 3 M H2SO4 solution or 0.01 M V(IV) in 3 M H2SO4 was used as the anolyte,
and 3 M H2SO4 solution or 0.01 M V(III) in 3 M H2SO4 was used as the catholyte in two
chambers of the photoelectrochemical cell (PEC) separated by a Nafion 117 membrane.
49
Solar irradiation was created using an ozone-free solar simulator system
(Newport USA) coupled with an AM 1.5 global filter (Newport USA) and calibrated using a
standard photodiode (Newport USA). The electrochemical impedance spectroscopy (EIS)
measurements in this study were performed on the electrochemical cell using either a
TiO2 or a hybrid electrode (24 wt% WO3) under dark/AM1.5 illumination conditions. All
data were recorded at open-circuit voltage (OCV) over a frequency range from 10 mHz to
2 MHz with an amplitude of 10 mV.
4.3 Zero-resistance Ammetry Study
We previously investigated the ability of WO3 as an electron reservoir to store
solar energy in an all-vanadium photoelectrochemical cell.[36, 43] Such energy storage
ability of WO3 is believed to be enabled through certain electrochemical conditions.
However, our previous work only focused on revealing the photoelectrochemical behavior
of WO3 and WO3/TiO2 hybrid electrodes under light for short period of time (up to 3 min)
and therefore may not reflect long-term behavior of the cell. In addition, it is unclear if the
energy storage behavior of the system is reversible. To this end, prolonged
photoelectrochemical study (up to almost 4 hrs) was conducted by ZRA to investigate the
overall cell performance in terms of electron storage using two distinct electrolytes.
50
Figure 4-4 Photocurrent measurements of PECs with different electrodes in 3 M H2SO4
electrolytes under alternative dark/illumination conditions. The inset graph shows the
photocurrent of different electrodes from 5 to 10 min under illumination.
Figure 4-4 depicts the photocurrent of different electrodes in 3 M H2SO4
electrolyte under alternative dark/illumination conditions and the inset graph magnifies
the photoresponse from 5 to 10 min under illumination. Generally, all electrodes show
relatively small photocurrent under AM 1.5 illumination, which is an indication of sluggish
reaction kinetics of H2 and O2 evolution. Compared to the TiO2 electrode, the hybrid
electrodes show even inferior photocurrent within almost the entire test window, though it
has been observed that increasing loading of WO3 in the hybrid electrodes resulted in a
red shift of absorption shoulder occurred near 450 nm.[36]
However, the inset graph indicates higher photocurrent for the hybrid electrode
with 12 wt% of WO3 than that of TiO2 electrode during the first 5 min illumination, which is
0 30 60 90 120 150 180 210
0.00
0.02
0.04
0.06
Da
rk
I, m
A
Time, min
AM
1.5
Time, min5 6 7 8 9 10
0.00
0.02
0.04
0.06
0.08
0.10
0.12 TiO
2
Hybrid, 1 wt%
Hybrid, 6 wt%
Hybrid, 12 wt%
Hybrid, 24 wt%
I, m
A
51
consistent with our previous findings.[43] In addition, the photocurrents produced by the
hybrid electrodes appear to reach saturation/stabilization eventually when the fraction of
WO3 is no less than 6 wt% in despite of illumination.
On the other hand, the hybrid electrodes give much more appreciable dark
current (at least one order of magnitude higher) than TiO2 electrode. Unlike the negligible
dark current by TiO2, the one by the hybrid electrodes remains almost without any decay
through each dark period (5 min). These findings achieved in prolonged test seem
contradictive to our previous results observed in short-time tests, but can be well
explained through the reversible intercalation/de-intercalation of electrons and H+ ions
into/out of WO3 to form hydrogen tungsten bronze, HxWO3. This is clearly seen from the
appearance of all hybrid electrodes (Figure 4-5) before and after the
photoelectrochemical experiments. The deeply colored blue-black hydrogen tungsten
bronze appeared on all hybrid electrodes after AM 1.5 irradiation except the one with 1
wt% WO3.
Figure 4-5 Appearance of hybrid WO3/TiO2 electrodes with WO3 loadings of (A) 1 wt%,
(B) 6 wt%, (C) 12 wt%, and (D) 24 wt% before and after ZRA experiment in 3 M H2SO4
electrolytes.
52
This observation coincides strongly with the photocurrent profiles of all hybrid
electrodes in Figure 4-4. When light was shed on the hybrid electrode, the
photogenerated electrons, apart from recombining with holes at the semiconductor/liquid
interface, have higher tendency to react with WO3 along with H+ ions to form HxWO3.
These HxWO3, scattered/distributed across the WO3/TiO2 matrix, are highly light-
reflecting due to their metallic or quasi-metallic nature[45-47] and believed to act as a
hurdle to electron transport and thus the photocurrent was reduced. This is especially
clear when the photocurrent collected by the hybrid electrodes was examined. When
WO3 content in the hybrid electrode is less dominant (e.g., 1 wt%), the photocurrent is
only slightly mitigated even under long-term illumination test. However, the photocurrent
was reduced to saturation/stabilization once the WO3 loading is more than 1 wt%
regardless of dark/illumination conditions. The noticeable dark currents from the hybrid
electrodes are due to the released electrons from decomposition reaction of hydrogen
tungsten bronze. However, the reaction kinetics of such electron release is believed to be
very sluggish in pure acid, resulting in unchangeable dark currents for the hybrid
electrodes especially with high WO3 loading.
Material characterization such as Raman spectroscopy and XRD was performed
on the hybrid electrode (24 wt%) before and after ZRA experiments to confirm the
formation of HxWO3. Raman spectra does not reveal observable difference on the
electrode whereas XRD results (Figure 4-6) on the other hand, show structural changes
of the hybrid electrode (24 wt%) after the ZRA tests.
53
20 30 40 50 60 70 80
2
Anatase TiO2
Rutile TiO2
Mono WO3
Before
After
(a)
22.5 23.0 23.5 24.0 24.5 25.0 2
Before
After
(b)
Figure 4-6 XRD spectra of the hybrid electrode (24 wt% WO3) in the range of (a) 20 – 80°
(2θ) and (b) 22.5 - 25° (2θ) showing WO3 characteristic peaks before and after ZRA
experiments in 3 M pure acid electrolyte.
In Figure 4-6A, only peaks of anatase TiO2 (JCPDS #21-1272), rutile TiO2
(JCPDS #76-1940), and monoclinic WO3 (JCPDS #83-0950) were found in the sample
before ZRA experiment. However, the crystal structure of WO3 in the sample changed
54
from monoclinic to cubic perovskite[48, 49] after intercalation of hydrogen ions into WO3
lattice. The major structural change appears mostly on three characteristic peaks of
monoclinic WO3 from 22.5° to 25° (Figure 4-6B) though other peaks remain the same
after the formation of HxWO3. It is clear that not only three major WO3 characteristic
peaks disappear, but also two new peaks emerge in different positions. Dickens[48]
reported that when hydrogen ions are inserted into WO3 lattice, all the corner-shared
WO6 octahedra are tilted relative to the orientation that is expected in a perovskite
structure. Thus, it is believed that such WO6 octahedra tilt is responsible for the observed
structural change of the hybrid electrode before and after the formation of HxWO3 in this
study.
Figure 4-7 Photocurrent measurements of PECs with different electrodes in 0.01 M all-
vanadium electrolytes under alternative dark/illumination conditions.
The photocurrents collected from various electrodes in Figure 4-7 when 0.01 M
all-vanadium electrolytes were used, are significantly different from what is shown in
Figure 4-4. The TiO2 electrode demonstrated distinctly different photoelectrochemical
TiO2
Hybrid, 1 wt%
Hybrid, 6 wt%
Hybrid, 12 wt%
Hybrid, 24 wt%
0 30 60 90 120
0.00
0.05
0.10
0.15
0.20
I, m
A
Time, min
Da
rk
AM
1.5
55
behavior than the hybrid ones. All hybrid electrodes show much higher photocurrent in
all-vanadium electrolyte compared to those in pure acid electrolyte. The improvement on
photocurrent is attributed to mitigated charge carrier recombination and fast reaction
kinetics of vanadium redox species. In addition, the amount of WO3 in the hybrid
electrode was discovered to play an important role affecting the photoelectrochemical
behavior of electrodes. When WO3 content in hybrid electrode was either inadequate or
abundant, the photocurrent was mitigated compared to that of TiO2 and only medium
WO3 loading (12 wt%) in the hybrid electrode gave optimum photocurrent. This
observation can be ascribed to the competition among vanadium redox reactions, charge
carrier recombination, and tungsten bronze formation reaction at the semiconductor/liquid
interface. When vanadium redox species were involved in the electrolyte, the
photoelectrons preferentially participated in redox reactions in addition to those in forming
hydrogen tungsten bronze, due to fast reaction kinetics of vanadium species and the
narrow-band-gap WO3. It is deduced that the charge carrier recombination reaction
dominates over the other reactions when WO3 loading is insignificant; whereas WO3 may
induce charge trapping effects in the bulk[50] and form more hydrogen tungsten bronze
when its loading is too high. An optimum amount of WO3 in the hybrid electrodes serves
as a mediator for effective charge carrier separation by minimizing the recombination
losses with the assistance of vanadium redox species during illumination.
56
5 10 15 20 25 30 35 400.000
0.005
0.010
0.015
0.020
AM
1.5
Dark
I, m
A
Time, min
TiO2
Hybrid, 24 wt%
AM
1.5
Figure 4-8 Significant electron storage capacity of the hybrid electrode (24 wt% WO3)
compared to the TiO2 electrode in 0.01 M all-vanadium electrolytes after the first 5 min
illumination. The graph was rearranged from Figure 4-7.
On the other hand, the ability of the hybrid electrodes to store photogenerated
electrons compared to TiO2 is conspicuously manifested in all-vanadium electrolytes by
their dark current shown in Figure 4-7 and Figure 4-8. Note that the test protocol shown
in the graph was slightly adjusted from that in Figure 4-4 to better exhibit the significant
dark currents of the hybrid electrodes under illumination according to our preliminary
experiments. In Figure 4-7, the hybrid electrodes were found to release electrons under
dark condition for prolonged periods of time and the dark current was almost linearly
proportional to the amount of WO3 in the hybrid electrodes.
The hybrid electrode with 24 wt% WO3 exhibits substantial electron storage
capacity under dark (estimated to be 0.299 mAh/g or 1.078 C/g) in the form of prolonged
57
electron release up to 30 min (Figure 4-8) with only 5-min illumination. The released
electrons from the hybrid electrodes under dark are believed due to decomposition of
hydrogen tungsten bronze formed during illumination and they can be used to continue
reducing the vanadium redox species even under dark. In other words, the
photoelectrochemical cell could keep being charged even under dark condition and this
may open a new perspective for photoelectrochemical solar energy conversion and
storage.
4.4 Electrochemical Impedance Spectroscopy Study
As exhibited in Figure 4-4 and Figure 4-7, hydrogen tungsten bronze formed
under illumination is believed to block the semiconductor/electrolyte interface and thus
compromise the photocurrent. It is crucial then to study the charge transfer at the
interface between the photoelectrode and electrolyte before, during and after the
illumination to understand its implication to photoelectrochemical kinetics.
Electrochemical impedance spectroscopy (EIS) is one of the most powerful
techniques for investigating charge transfer processes in an electrochemical device. To
this end, EIS measurement was conducted at OCV of the cells on different
photoelectrodes under dark/illumination conditions. Figure 4-9 and Figure 4-10 represent
the Nyquist plots of the TiO2 and hybrid (24 wt% WO3) electrodes, respectively, in pure
acid or 0.01 M all-vanadium electrolytes under dark and illumination conditions. As shown
in Figure 4-9 and Figure 4-10, TiO2 and hybrid electrodes show distinct characteristics of
impedance in two electrolytes albeit their spectra are all composed of similar arcs/partial
arcs. Generally, an arc in the Nyquist plot represents existence of an electrochemical
interface and the appearance of plural arcs indicates different time constants of
electrochemical interfaces for electron transport in the electrochemical system.
58
Figure 4-9 Nyquist plots of PECs in (A) 3 M H2SO4 and (B) 0.01 M all-vanadium with a
TiO2 photoelectrode. The inset graph in Figure 4-9B represents Nyquist plots of the cell in
all tested frequency.
When only 3 M H2SO4 electrolyte was used (Figure 4-9A), the high-frequency arc
and an even bigger arc (only a portion) at low frequency, according to consensus in
0 5 10 15 20 250
5
10
15
20
Zim
,
Zre
,
Dark, onset
AM1.5, 1 min
AM1.5, 30 min
AM1.5, 240 min
Dark, 1 min
Dark, 30 min
Dark, end
(A)
Dark, onset
AM1.5, 1 min
AM1.5, 30 min
AM1.5, 240 min
Dark, 1 min
Dark, 30 min
Dark, end
0 5 10 15 20 250
5
10
15
20
0 500 1000 1500 20000
500
1000
Zim
,
Zre
,
(B)
Zim
,
Zre
,
59
literature,[51, 52] are believed to correspond to electron transport resistance and
interfacial capacitance at Pt/electrolyte and TiO2/electrolyte interface, respectively. The
dominantly large diameter of the low-frequency arc indicates huge charge transfer
resistance and thus sluggish reaction kinetics[53] regardless of illumination. This finding
corroborates with the observation in Figure 4-4, indicating water splitting by TiO2 is
intrinsically sluggish. Note that the diameter of the small arc at high frequency is
independent of dark/illumination conditions, proving that no change occurred on Pt
electrode itself.
However, when vanadium redox species were involved in the electrolyte (Figure
4-9B) and new interfaces were created, big difference is seen on impedance spectra
compared to Figure 4-9A. One more arc with much larger diameter emerged at mid
frequency and it represents the charge transfer resistance and interfacial capacitance at
TiO2/vanadium(IV) redox interface. The new arc at mid frequency dominates much more
in the whole impedance spectra than electron transport at high frequency and Warburg
diffusion resistance at low frequency. Though not fully understood, this result indicates
that vanadium redox plays a vital role at TiO2/vanadium(IV) redox and Pt/vanadium(III)
redox interfaces in terms of charge transfer resistance and electron transport resistance.
In addition, the new interfaces created by vanadium redox species seem to have
little influence on electron transport resistance at the counter electrode as the resistance
values (see magnified inset graph in Figure 4-9B are very similar to their counterparts in
Figure 4-9A regardless of dark/illumination conditions.
60
Figure 4-10 Nyquist impedance spectra of PECs in (A) 3 M H2SO4 electrolytes and (B)
0.01 M all-vanadium electrolytes with a hybrid (24 wt% WO3) photoelectrode. The inset
graph in Figure 4-10B represents Nyquist plots of the cell under various conditions in all
tested frequencies.
0 10 20 30 40 50 60 700
10
20
30
40
Zim
,
Zre
,
(A)
Dark, onset
AM1.5, 1 min
AM1.5, 30 min
AM1.5, 240 min
Dark, 1 min
Dark, 30 min
Dark, end
0 10 20 30 40 50 60 700
10
20
30
40
0 500 1000 15000
200
400
600
800
Zim
,
Zre
,
Zim
,
Zre
,
Dark, onset
AM1.5, 1 min
AM1.5, 30 min
AM1.5, 240 min
Dark, 1 min
Dark, 30 min
Dark, end
(B)
61
However, the charge transfer resistance with vanadium redox species was
considerably reduced up to three times under AM 1.5 illumination. This is due to
facilitated electron generation and transport by fast reaction kinetics of vanadium redox,
which is in agreement with Figure 4-7. As indicated in our previous study,[43] the
vanadium redox also shifts the dynamic balance between charge carrier recombination
and redox reaction toward the latter, which is verified in Figure 4-9B. The electron
transport resistance (diameter of the mid-frequency arc) after illumination for 1 min (red
line) was greatly reduced and quickly reached stabilization in 30 min (blue line). The EIS
spectra remained unchanged even up to 4 hrs (green line). Besides, when the light was
turned off, the electron transport resistance was promptly increased and reverted back to
its original value as before the test. Comparison between Figure 4-9A and Figure 4-9B
clearly reveals the quick charge transfer at the semiconductor/electrolyte interface when
vanadium redox species were employed.
Figure 4-10 depicts the Nyquist plots of the hybrid (24 wt% WO3) electrode in the
photoelectrochemical cell using two electrolytes. They have similar characteristic
arcs/diffusion components as those of TiO2 in Figure 4-9, but with some obvious
differences. As suspected from Figure 4-4, hydrogen tungsten bronze formed under
illumination by WO3 reacting with photogenerated electrons and H+ ions in the electrolyte
is realized as a hurdle to electron transport at the semiconductor/electrolyte interface.
Such speculation is confirmed from the impedance spectra shown in Figure 4-10. The
electron transport resistance corresponding to the Pt/electrolyte interface (i.e., diameter
of the high-frequency arc) in both electrolytes was nearly doubled under illumination and
this increase was strictly dependent on time.
62
Figure 4-11 Appearance of hybrid WO3/TiO2 electrodes with WO3 loadings of (A) 1 wt%,
(B) 6 wt%, (C) 12 wt%, and (D) 24 wt% before and after ZRA experiment in 0.01 M all-
vanadium electrolytes.
Though the nature of the Pt/electrolyte interface at the counter electrode
remained unchanged; however, the charge transfer kinetics might be impeded under
illumination because the incoming electrons from the photoelectrode are blocked by the
formation of HxWO3. It is seen in Figure 4-10B that an increase in electron transport
resistance, as a result of HxWO3 formation, appeared delayed in all-vanadium
electrolytes. Besides, the electron transport resistance after long-term illumination (4 hrs)
was about 5 Ω lower than that in 3 M H2SO4 electrolytes. Such results are consistent with
the observed saturated photocurrent of the hybrid (24 wt%) electrode in Figure 4-4, and
the electrode appearance change in Figure 4-5 and Figure 4-11. It is believed that fast
reactions of vanadium(IV) redox species at the photoelectrode surface, which quickly
consume the photogenerated electrons, leave only a small fraction of those to participate
in the formation of HxWO3. This significantly delays growth of HxWO3 on the
photoelectrode and results in slow changes of both high-frequency and mid-frequency
arcs in Figure 4-10B. The above analysis was confirmed by the unchanged electrode
63
appearance (Figure 4-11) for different WO3 loadings in 0.01 M all-vanadium electrolyte
before and after the test.
4.5 Electron Storage Mechanism
To better understand the observed EIS spectra of the hybrid electrode in two
electrolyte systems under different dark/illumination conditions, a model based upon the
relevant electronic states of different components and standard electrochemical potential
of different vanadium redox species is proposed in Figure 4-12.
Figure 4-12 Schematic representation of proposed electron storage mechanism and
charge transfer pathways for the hybrid WO3/TiO2 electrode in the all-vanadium
photoelectrochemical storage cell.
64
This model also helps further elucidate the electron storage mechanism and
charge transfer pathway for the hybrid electrode. In this model, to avoid unnecessary
ambiguity, the bandgap energy of WO3 is designated as 2.6 eV and the conduction band
(CB) edge is about 0.4 V more positive than that of TiO2 in the electrochemical scale as
is typically reported in literatures.[54-59] Although controversy remains in the academic
society that the bandgap energy of WO3 spans from 2.6 to 3.3 eV. However, these values
mainly differ in the position/maximum energy level of valence band (VB) and thus shall
have little impact in our discussion regarding the electron storage mechanism for the
hybrid electrode and charge carrier transfer processes.
As illustrated in Figure 4-12, the photogenerated electrons and holes in TiO2
need to travel ―downhill‖ and ―uphill‖ to the CB and VB of WO3 respectively because of
their relative band positions. The more positive CB and more negative VB of WO3 are
deemed as the driving force for such charge carrier migration. During this process,
different reaction scenarios emerge depending on the electrolyte used. When pure acid
was employed as the electrolyte, photogenerated holes from TiO2 and WO3 oxidize H2O
to result in O2 evolution while the photoelectrons react with WO3 and H+ ions to form
hydrogen tungsten bronze. This is why the blue-black coloration appeared on the
electrodes after illumination in pure acid electrolyte. Note that such photochromism highly
depends on the amount of WO3 that participates the reaction and electron transfer
kinetics. When WO3 loading in hybrid electrode is insignificant (e.g. 1 wt%), the
photochromism is insignificant (Figure 4-5) as little difference was found after long-term
illumination up to 4 hrs by examining the electrode appearance. However, noticeable
dark current observed after switching off the light clearly indicates the formation of
hydrogen tungsten bronze and its slow decomposition to release electrons. As for the
hybrid electrodes with higher WO3 loading, blue-black coloration (Figure 4-5) was
65
observed after the same test and the photocurrent (Figure 4-4) during the test quickly
reached to saturation/stabilization regardless of dark/illumination condition. These
observations confirm that (i) hydrogen tungsten bronze is preferably formed out of hybrid
electrodes under AM 1.5 illumination, (ii) it renders as the barricade to prevent electron
transfer, (iii) the decomposition is very slow as revealed in Fujishma’s work.[39]
When vanadium redox species were involved in the electrolyte, photogenerated
charge carriers, have a distinct pathway at the semiconductor/liquid interface. Similar as
in the previous scenario, photogenerated holes from TiO2 will still travel ―uphill‖ to the
WO3 VB and join their counterparts there. Instead of evolving O2, these holes are inclined
to oxidize vanadium(IV) ions to vanadium(V) ions due to more negative electrochemical
potential and faster reaction kinetics of V4+
/V5+
redox[36, 43] compared to oxygen
evolution reaction. On the other hand, the photogenerated electrons from TiO2 are
speculated to mainly/completely react with vanadium(III) ions in the catholyte and reduce
them to vanadium(II) ions when they travel ―downhill‖ to the conduction band of WO3 as a
result of fast electrochemical kinetics of V3+
/V2+
redox. The remaining photogenerated
electrons from TiO2, if any, along with their counterparts in WO3 will react with WO3 to
form hydrogen tungsten bronze. This process is reverted under dark upon demand
releasing the stored electrons. This finding is confirmed by the prolonged dark current
(Figure 4-7 and Figure 4-8) and EIS measurement (Figure 4.10B) as the resistance was
increased after long-term illumination, which indicates the existence of hydrogen tungsten
bronze.
Meanwhile, the fast reaction kinetics of V3+
/V2+
redox is believed to prevent TiO2
photoelectrons from recombination, thus providing much higher photocurrent than that is
exhibited in pure acid electrolyte. In other words, hybrid electrode is capable of
revitalizing photocurrent in all-vanadium electrolytes under illumination and releasing
66
electron energy stored in hydrogen tungsten bronze under dark simultaneously. Note that
the magnitude of photocurrent revitalized and the electron energy released strongly
depend on the amount of WO3 in hybrid electrode and redox reaction kinetics. Most
important, the electron storage capability of the hybrid electrode when coupled with the
all-vanadium electrolytes, in comparison with pure acid electrolytes, potentially offer
great reversibility (Figure 4-11), long-term electron storage (up to 4 hrs in Figure 4-10),
and significant improvement in photocurrent (Figure 4-5). Since what has been
demonstrated here in this work only employed unmodified semiconductors, it is
envisioned that further improvement can be brought in by future optimization to enable
practically meaningful continuous solar energy conversion and storage using advanced
semiconductors in tandem structure.
4.6 Conclusion
The energy storage ability of WO3 was investigated using WO3/TiO2 hybrid
photoelectrodes in an all-vanadium photoelectrochemical storage cell under either dark
or illumination conditions. ZRA and EIS were used to probe photoelectrochemical and
electrochemical behavior of the cell with respect to photoelectrochemical solar energy
conversion and storage.
Results revealed mitigated photocurrent in pure acid electrolyte but enhanced
photoresponse of the hybrid electrodes in all-vanadium electrolytes under long-term
illumination. Without applying the bias, hydrogen tungsten bronze was formed in both
pure acid and all-vanadium electrolytes, and this finding was corroborated by EIS results
and material characterization results. In pure acid, the formed hydrogen tungsten bronze
acts as the hurdle to prevent electron transport and thus mitigates the photocurrent,
whereas all-vanadium electrolytes are beneficial to revitalize photocurrent of the hybrid
electrodes under illumination and realize energy storage ability of WO3 under dark.
67
However, a complicated interplay was found to exist between achieved
photocurrent and released dark current by using different WO3 loadings in the hybrid
photoelectrode. A model is then proposed to elaborate the observed experimental results
and the charge carrier transfer pathways. The vanadium redox species, due to their fast
electrochemical kinetics, not only enhance the photocurrents under illumination through
mitigating formation of HxWO3, but also help release reversely stored electrons under
dark.
68
Chapter 5
Electrolyte Study of All-Vanadium PEC
5.1 Introduction
Previously, we studied the photoresponse of typical n-type photoelectrodes such
as TiO2 and WO3/TiO2 with the assistance of vanadium redox species. The initial results
show significantly enhanced photoelectrochemical performance (15 times higher
IPCE[44]) of the PEC even with an non-optimized system when vanadium redox species
were added in the electrolyte.
To realize even better photocatalytic property of the PEC, optimization on cell
components, e.g., electrode, membrane and electrolyte is necessary. However, the state-
of-the-art electrode and membrane normally require delicate material synthesis or
fabrication and laborious procedures and thus time-consuming. Replacing the electrolyte
instead deems it a very fast, convenient and cost-effective approach to achieve the goal.
Therefore, seeking for electrolytes especially acid-based ones with high solution
conductivity, superior electrochemical activity, chemical stability and environment non-
toxicity prioritizes our tasks of cell optimization for better photoelectrochemical
performance.
Methanesulfonc acid (MSA) has been widely used as a commercial electrolyte
standard in the past three decades to replace the previous industrial standard, fluoroboric
acid in many electrochemical processes especially those involving lead and tin. This is
due to its diverse physical and chemical properties such as good thermal stability, high
water miscibility and metal salts solubility, low relative toxicity and high conductivity. It is
already reported that the conductivity of 1.0 mol·L−1
MSA aqueous solution is 0.30
S·cm−1
, comparable to that of hydrochloric acid and sulfuric acid (0.35 S·cm−1
and 0.44
S·cm−1
, respectively), while posing a lower risk of corrosion compared with other mineral
69
acids. In addition, MSA aqueous solutions exposed to open atmospheric conditions
display a unique stabilization of metal ions in their lower valence states, or, stated
differently, MSA solutions allow for a unique resistance to the oxidation of metal ions to
their higher valence states.
Due to these distinguished properties, using MSA as the supporting electrolyte to
improve electrochemical properties has become a hot topic in flow battery research
recent years.[60-64] These preliminary electrochemical studies all reveal its more
pronounced metal salt solubility, redox reaction reversibility, reaction kinetics and cell
efficiencies compared to conventional sulfuric acid. More recently, MSA has been
employed as the electrolyte in PEC study by a group of Poland researchers[65] and their
results clearly indicate long-term stability and more admirable photoelectrochemical
performance achieved by MSA than H2SO4 and HClO4 on a WO3 nanostructured
photoelectrode under the same acid concentration.
In this chapter, we investigated long-term chemical stability, bulk conductivity,
electrochemical and photoelectrochemical performance of our all-vanadium PEC using
MSA as the supporting electrolyte by four-point electrical conductivity measurement, CV,
ZRA and EIS. The results, compared to those achieved under the same conditions using
H2SO4 as the supporting electrolyte show its great potential to further enhance
photoelectrochemical performance of all-vanadium PEC.
5.2 Experiments
TiO2 photoelectrode with active area of 1.61 cm2 were fabricated and used
throughout the experiment. To fabricate TiO2 electrode, 1.00 g Degussa P25 TiO2 (VP
AEROPERL® by Evonik), 2.50 g α-terpineol (laboratory grade, Fisher Scientific USA)
were mixed under constant stirring at 80°C for 1 hr to obtain a uniform TiO2 slurry. Then
the slurry was deposited on a pre-cut square-shaped fluorine doped tin oxide (FTO)
70
(Pilkington USA) using a doctor blade. The FTO substrate was pre-washed with acetone
(99.7%, Fisher Scientific USA), methanol (99.8%, Fisher Scientific USA), and deionized
(DI) water, before being blow-dried and then further dried in an oven at 120°C for 1 hr.
The obtained coating was subsequently calcined with air flow at 500°C for 90 min. Since
the electrode fabrication process is very reproductive, material characterization results
are thus referred to previous chapters.
Six types of electrolytes, 3 M H2SO4 or MSA, 0.01 M vanadium(IV, VO2+
) in 3 M
H2SO4 or MSA, and 0.01 M vanadium(III, V3+
) in 3 M H2SO4 or MSA were used in the
experiments. The H2SO4-based electrolytes were prepared by dissolving H2SO4 (96.6%,
J.T. Baker USA) in DI water with or without vanadium(IV) sulfate oxide hydrate
(VOSO4·xH2O) (99.9%, Alfa Aesar USA) while the MSA-based electrolytes were
prepared by dissolving MSA (98%, Alfa Aesar USA) in DI water with or without
vanadium(IV) sulfate oxide hydrate (VOSO4·xH2O) (99.9%, Alfa Aesar USA). The number
of water in VOSO4·xH2O was determined by thermogravimetric analysis. The prepared
vanadium(IV)-sulfuric acid and vanadium(IV)-MSA solution both appeared light blue. The
0.01 M vanadium(III) electrolyte was obtained by electrochemically reducing the prepared
vanadium(IV)-H2SO4 or vanadium(IV)-MSA solution in an electrolytic cell at constant
current density of 3 mA/cm2 using a potentiostat (Princeton Applied Research, PARSTAT
2273) until the potential reached 1.6 V. During the reduction, the electrolyte was
protected by N2 to prevent oxidation of the vanadium(III) species. The obtained
vanadium(III)-sulfuric acid and vanadium(IV)-MSA solution appeared light green as well.
The photoelectrochemical property of TiO2 electrode was studied by CV, ZRA
and EIS in a two-chamber, three-electrode electrochemical cell, where the
photoelectrode served as the working electrode (WE), a platinum mesh and Ag/AgCl
electrode served as the counter electrode (CE) and reference electrodes (RE),
71
respectively. Details of the experimental setup have been described in previous chapters.
Tests using zero-resistance ammetry (ZRA) were conducted with either 3 M pure acid or
0.01 M vanadium-pure acid electrolyte under alternate dark and AM 1.5 conditions. In a
typical experiment, 3 M H2SO4 or MSA solution or 0.01 M V(IV) in 3 M H2SO4 or MSA
was used as the anolyte, and 3 M H2SO4 or MSA solution or 0.01 M V(III) in 3 M H2SO4
or MSA was used as the catholyte in two chambers of the PEC separated by a Nafion
117 membrane.
Solar irradiation was created using an ozone-free solar simulator system
(Newport USA) coupled with an AM 1.5 global filter (Newport USA) and calibrated using a
standard photodiode (Newport USA). The electrochemical impedance spectroscopy (EIS)
measurements in this study were performed on the electrochemical cell using a TiO2
electrode under dark/AM 1.5 illumination conditions. All data were recorded at open-
circuit voltage (OCV) over a frequency range from 1 mHz to 2 MHz with an amplitude of
10 mV.
5.3 Cyclic Voltammetry and Zero-resistance Ammetry Study
Cyclic voltammetry was employed to investigate the chemical stability of
methanesulfonic acid while the PEC was in operation. Figure 5-1 shows the cyclic
voltammogram of the PEC using pure MSA as the electrolyte in conjunction with a TiO2
photoelectrode under a large scan range. Fifty scans were conducted on the PEC under
alternate dark and AM 1.5 illumination conditions. As shown in Figure 5-1, TiO2 shows a
typical n-type semiconductor photoelectrochemical behavior in MSA electrolyte. Under
the scan range, no other oxidation/reduction peak other than the apparent one near -0.11
V is observed on anodic scans. This peak is ascribed to charge compensating proton
adsorption or intercalation reaction on TiO2 in acidic aqueous electrolyte[66-68] and is
described schematically by the brutto reaction:
72
𝑇𝑖𝐼𝑉𝑂2 + 𝑒− +𝐻+ ↔ 𝑇𝑖𝐼𝐼𝐼(𝑂)(𝑂𝐻)………………5.1
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
-4
-3
-2
-1
0
1
2
3
4
I, m
A
Potential vs Ag/AgCl, V
1st scan, Dark
21st scan, AM1.5
30th scan, AM1.5
50th scan, Dark
Figure 5-1 Cyclic voltammetry of a PEC using pure MSA as the electrolyte and TiO2 as
the photoelectrode with scan range of -0.5 to 2.1V under dark and AM 1.5 illumination.
The scan rate was 20 mV/s.
In addition, above 2.0 V, the current starts to increase rapidly due to sufficient
overpotential being provided to split water with fast reaction rate, either electrolytically
(under dark) or photoelectrolytically (under AM 1.5). The CV results shown in Figure 5-1
confirm that MSA is a chemical stable electrolyte to participate in electrochemical and
photoelectrochemical reactions.
73
-0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
I, m
A
Potential vs Ag/AgCl, V
3 M H2SO
4, Dark
3 M H2SO
4, AM 1.5
3 M MSA, Dark
3 M MSA, AM 1.5
Figure 5-2 Linear sweep voltammetry of a PEC using TiO2 as the photoelectrode in 3 M
H2SO4 and 3 M MSA electrolytes under dark and AM 1.5 illumination.
Figure 5-2 compares photoelectrochemical performance of a TiO2 photoelectrode
in 3 M H2SO4 and 3 M MSA electrolytes under dark and AM 1.5 illumination by linear
sweep voltammetry. It is shown that MSA improves photocurrent significantly than H2SO4
at the same concentration on the TiO2 photoelectrode and a four-fold improvement in
photocurrent was achieved. Considering the fact that the concentration of MSA used in
our experiments are very high, i.e., negative pH value of the electrolyte, the above
preliminary results render MSA a very promising electrolyte compared to conventional
H2SO4 for photoelectrochemical solar energy conversion and storage.
However, these results rely on cyclic voltammetry that requires an external bias
applied to the cell. In order to further investigate photoelectrochemical performance of
MSA with more practical meaning, ZRA method (no applied external bias) was conducted
74
on TiO2 photoelectrode in various electrolytes with or without the support of vanadium
redox, and the results are shown in Figure 5-3. As seen in Figure 5-3, pure MSA is
chemically stable upon illumination and shows clearly enhanced photocurrent (5 times
higher) than pure H2SO4 over the entire test window under AM 1.5 illumination. Albeit
current spikes, attributed to surface trap states of TiO2 were observed at the beginning
upon each illumination, the photocurrent reaches equilibration eventually after short
period of time. This result is in alignment with above CV findings and it further justifies
MSA being a promising electrolyte for the photoelectrochemical solar energy conversion
and storage. What seems interesting in Figure 5-3 is that and MSA gives higher
photocurrent on a TiO2 photoelectrode even than 0.01 M V in 3 M H2SO4 electrolyte,
which might indicate higher diffusion coefficient of electroactive species in the former
compared to the latter. It is suspected that 0.01 M V in 3 M MSA would show even more
enhanced photoelectrochemical performance when vanadium redox is involved in the
electrolyte. Such speculation is confirmed by green line in Figure 5-3. When vanadium is
involved in the electrolyte, the photocurrent of TiO2 in V-MSA electrolyte is boosted more
than 6 times than that in V-H2SO4 electrolyte under the same concentration after it is
stabilized. Note that there are two different ZRA profiles using the same TiO2
photoelectrodes emerging in various electrolytes. The ZRA profile using H2SO4, V-H2SO4
and V-MSA resemble with each other, showing gradually increased photocurrent till
equilibration upon illumination while ZRA profile using MSA shows gradually dropped
photocurrent till equilibration upon illumination. In our preliminary experiments, we
discovered that ZRA profile using MSA also shows a gradually increased photocurrent at
the beginning upon illumination. However, the increased photocurrent starts to drop
abruptly till equilibration after certain period of time. Thus it is more scientific to use ZRA
profile of the electrolyte after the photocurrent reaches to equilibration. Meanwhile, the
75
ZRA profiles using MSA and V-MSA electrolytes both show residual dark current in
Figure 5-3. Such dark current is believed to be due to uncompensated charge carrier
adsorption at the semiconductor/liquid interface immediately after light off. These
uncompensated charge carriers can be eliminated by discharging the PEC under dark for
long period of time according to our experiments (not shown here).
0 40 80 120 160 200 240
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Da
rk
AM
1.5
I, m
A
Time, s
3 M H2SO
4
0.01 M V in 3 M H2SO
4
3 M MSA
0.01 M V in 3 M MSA
Da
rk
Da
rk
AM
1.5
Da
rk
AM
1.5
Da
rk
AM
1.5
Da
rk
AM
1.5
Da
rk
AM
1.5
Figure 5-3 Photocurrents of TiO2 photoelectrode using ZRA method in various
electrolytes under alternate dark and AM 1.5 illumination conditions.
As demonstrated by both CV and ZRA experiments, MSA is much more superior
electrolyte compared to H2SO4 in terms of both photoelectrochemical performance and
chemical stability. However, the root cause of its better performance over H2SO4 remains
unknown to us, so it is crucial to investigate electrochemical and photoelectrochemical
reaction kinetics and mechanism to unfold the fundamentals behind.
76
5.4 Bulk Ionic Conductivity Study
We started first by studying bulk ionic conductivity of the electrolyte using
electrochemical impedance spectroscopy and four-probe electrical conductivity
measurement. The Nyquist plot of a homemade test cell using various electrolytes are
shown in Figure 5-4. Four different electrolytes were tested and all samples show very
similar semi-circle indicating the interface between the electrode and the electrolyte. The
interception of these semi-circles on Y=0 gives ohmic resistance of the cell.
0 2 4 6 8 10
0
1
2
3
4
5
Zim
,
Zre,
3 M H2SO
4
0.01 M V in 3 M H2SO
4
3 M MSA
0.01 M V in 3 M MSA
Figure 5-4 Nyquist plots of the test cell using various electrolytes.
The bulk ionic conductivity of various electrolytes is calculated according to the
following equation:
𝜎 =𝐿
𝑍𝑟𝑒 ∙𝐴………… …………..……….……..5.2
wherein:
σ: the ionic conductivity of the electrolyte
77
L: the overall length of four sensing probes
Zre: ohmic resistance of the cell
A: electrode area available for ionic conduction (width x thickness)
Table 5-1 shows calculated bulk ionic conductivity of various electrolytes using a
graphite electrode. All tested electrolytes have very similar ionic conductivity values when
they are compared. Compared to pure H2SO4, pure MSA has slightly lower bulk ionic
conductivity but within the same order of magnitude. This result is consistent with the
discovery revealed in the literature.[65, 69] Besides, the bulk ionic conductivity of both
electrolytes increases after vanadium ions were added in the solution, which is due to the
enhanced ionic strength due to additional vanadium ions in both acid.
Table 5-1 Calculated ionic conductivity of various electrolytes.
Electrolytes Ohmic Resistance (Ω) Conductivity (S/cm)
3 M H2SO4 0.9487 5.8972
0.01 M V in 3 M H2SO4 0.8149 6.8655
3 M MSA 1.1158 5.0141
0.01 M V in 3 M MSA 0.9621 5.8151
5.4 Electrochemical Impedance Spectroscopy Study
EIS was also employed to investigate reaction kinetics and mechanism on MSA
based electrolytes given our previous success.[70] Figure 5-5 represents (A) Nyquist plot
and (B) Bode plot of a TiO2 photoelectrode in various electrolytes under AM 1.5
illumination. As seen in Figure 5-5A, one semi-circle at high frequency and one arc/partial
arc at mid frequency were observed in all tested electrolytes. The semi-circle
corresponds to electron transport resistance and interfacial capacitance at Pt/electrolyte
while the arc/partial arc represents charge transfer resistance and interfacial capacitance
at TiO2/vanadium(IV) redox interface. It is clear that all electrolyte has little influence on
78
electron transport resistance and interfacial capacitance at Pt/electrolyte interface as they
all show very similar value near 50 Ω of the real resistance.
0 1000 2000 3000 4000 5000 6000
0
1000
2000
3000
4000
0 50 100 150 200
0
40
80
120
Zim
,
Zre,
3MH2SO
4
0.01MV-3MH2SO
4
3MMSA
0.01MV-3MMSA
Zim
,
Zre,
3 M H2SO
4
0.01 M V-3 M H2SO
4
3 M MSA
0.01 M V-3 M MSA
(A)
3 M H2SO4
0.01 M V-3 M H2SO4
3 M MSA
0.01 M V-3 M MSA
10-2
10-1
100
101
102
103
104
0
20
40
60
80
100
-Ph
ase
()
Frequency (Hz)
(B)
Figure 5-5 Electrochemical impedance spectroscopy of a PEC in the form of (A) Nyquist
plot and (B) Bode plot using TiO2 as the photoelectrode in various electrolytes under AM
1.5 illumination.
79
However, big difference is seen at the TiO2/electrolyte interface depending on the
selection of electrolyte. 3 M H2SO4 electrolyte shows the mid frequency arc with the
biggest diameter compared to other three even under AM 1.5 illumination, this is an
indicative of slow kinetics of water splitting reactions. After vanadium redox is involved in
the electrolyte, charge transfer resistance and interfacial capacitance at the
TiO2/electrolyte interface is reduced greatly due to fast reaction kinetics of vanadium
redox, similar as revealed in our previous work.[36, 43, 70] The same argument can be
applied to MSA electrolyte as well. Other than that, both MSA-based electrolytes exhibit
much smaller charge transfer resistance and interfacial capacitance than H2SO4-based
electrolytes regardless of vanadium redox participation at TiO2/electrolyte interface. The
0.01 M V in 3 M MSA electrolyte especially, displays approximate 5 time smaller
resistance compared to 0.01 M V in 3 M H2SO4. These results are in great agreement
with the CV and ZRA results, further proving that MSA holds a great potential as an
encouraging electrolyte in photoelectrochemical solar energy conversion and storage.
Figure 5-5B shows Bode plots of a PEC using TiO2 as the photoelectrode in
various electrolytes under AM 1.5 illumination, which is capable of providing lifetime
information of the photogenerated electrons in the reaction (calculated according to the
following equation and listed in Table 5-2)
𝜏𝑒 =1
2𝜋𝑓𝑚𝑎𝑥……………………....5.3
Wherein:
τe: the lifetime of electrons
fmax: the maximum frequency of the peak in the low frequency region
Clear difference from all electrolytes is seen in Figure 5-5B. Peaks given by two
MSA-based electrolytes in despite of vanadium redox, both shift to lower frequency
region by two orders of magnitude compared to their counterparts in H2SO4 electrolyte as
80
shown in Table 5-2. This indicates the photoelectron lifetime in MSA-based electrolyte is
significantly prolonged compared to H2SO4-based electrolytes. This result is in good
alignment with our previous CV and ZRA results, indicating a diminished charge carrier
recombination and much higher photocatalytic property because of the MSA electrolyte,
especially in the presence of vanadium redox species. However, vanadium redox seems
to play a negative role in TiO2 photoelectrode by reducing photoelectron lifetime of TiO2
photoelectrode and this is true for both H2SO4 and MSA acid as the supporting
electrolyte. The ratios of photoelectron lifetime in vanadium-acid electrolyte to it in pure
acid electrolyte under both H2SO4 and MSA circumstances are calculated to be 0.45 and
0.43, which is very close to unity, indicating the same electrochemical and/or
photoelectrochemical behavior of vanadium redox in two different acids.
Now one question that needs to be answered is why the vanadium-based
electrolyte show much higher photocurrent than pure acid electrolyte regardless of the
acid kind whereas its electron lifetime profile exhibits the opposite trend. This is attributed
to the quick charge carriers scavenging effect of vanadium redox species due to its fast
reaction kinetics. The ratios of electron lifetime in vanadium-acid electrolyte to it in pure
acid electrolyte regardless of the supporting electrolyte kind closing to a unity, as
indicated earlier, is a strong evidence to indicate the same electrochemical and/or
photoelectrochemical behavior of vanadium redox species and to support such
speculation. These significantly scavenged charge carriers contribute greatly to the
photocurrent. Therefore, although electron lifetime of TiO2 electrode is shortened vastly in
vanadium-based electrolyte compare to pure acid electrolyte, fast reaction kinetics of
vanadium species still outperforms such negative effect as indicated in all our previous
studies[36, 43, 70] and thus still rendering higher photocurrent as demonstrated in Figure
5-2 and Figure 5-3.
81
Table 5-2 The maximum frequency of the peak and calculated electron lifetime of a TiO2
photoelectrode in various electrolytes under AM 1.5 illumination.
Electrolytes fmax (Hz) τe (ms)
3 M H2SO4 1 159.2
0.01 M V in 3 M H2SO4 2.205 72.2
3 M MSA 0.0235 6775.9
0.01 M V in 3 M MSA 0.055 2895.2
5.5 Long-term Stability Study
Furthermore, the long-term chemical stability of our all-vanadium PEC was
studied by conducting photoelectrochemical experiments for prolonged period of time and
examining the electrode before and after the experiment. Figure 5-6 illustrates the long-
term photoelectrochemical behavior of our all-vanadium PEC using the TiO2
photoelectrode in 0.01 M V-3 M MSA electrolyte under AM 1.5 illumination for 60 hrs. The
cell shows very stable photocurrent throughout the whole test window although slight
fluctuation caused by intentional interruption of the light is seen. The photocurrent
retention remains 89.5% even after 60 hrs operation. The photocurrent loss may be due
to slight concentration loss of vanadium redox species in the photoelectrode during the
test. Besides, trivial photoelectrode physical destruction from FTO may also contribute to
the loss as slight delamination of TiO2 was observed after examining the surface of FTO
substrate. Judging by our electrode fabrication procedure, our photoelectrode is more a
coating rather than a thin film and when the solid is immersed in the liquid for as long as
60 hrs, partial delaminating occurs naturally as the liquid electrolyte tends to swell the
solid. This mild electrode physical destruction can be improved by more advanced thin
film fabrication techniques, such as electrochemical deposition, chemical vapor
deposition (CVD), pulsed laser deposition (PLD) or physical vapor deposition (PVD) etc.
82
0 6 12 18 24 30 36 42 48 54 60
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
AM1.5
53h44h35h24h14h
I, m
A
Time, h
7h
Dark
Figure 5-6 Long-term photoelectrochemical profile of all-vanadium PEC using the TiO2
photoelectrode immersed in 0.01 M V-3 M MSA electrolytes under AM 1.5 illumination for
60 hrs.
XRD was performed on TiO2 photoelectrode to see any crystal structure/phase
change or impurity associated with the photoelectrochemical test and the result is shown
in Figure 5-7. It is clearly seen that the crystal structure of TiO2 remains the same. Only
anatase and rutile phase TiO2 are observed and no impurity detectable is discovered
though the intensity of the electrode diminishes a little after long-term test. This is in
agreement with observed partial electrode delamination from FTO as the fewer amount of
TiO2 tends to give relatively lower intensity on XRD.
83
20 30 40 50 60 70 80
2
Anatase TiO2
Rutile TiO2
Before ZRA test
After ZRA test
Figure 5-7 XRD spectra of TiO2 photoelectrode before and after 60 hrs cell operation.
5.6 Efficiencies Measurement
So far, all our experiments have demonstrated improved photocatalytic property
for the all-vanadium PEC employing vanadium redox species in the electrolyte. However,
these measurements were conducted qualitatively and thus comprehensive study to
further quantify the electrode, electrolyte and the system is needed. Herein, characteristic
measurements of all-vanadium PEC such as Faradaic efficiency, incident photo-to-
current conversion efficiency (IPCE) of the PEC were implemented to reveal its
photoelectrochemical performance quantitatively.
Recently, UV-vis spectroscopy has been developed[72-75] to monitor state of
charge of all-vanadium redox batteries, such method was also utilized in our study to
84
monitor the electrolyte concentration change while the all-vanadium PEC was in
operation. Figure 5.8 shows the concentration change of vanadium(IV, VO2+
) ions at the
beginning, in the middle and the end of the 60 hrs photocharging process as
demonstrated in Figure 5-6. As the photocharging process proceeds, the vanadium(IV)
ions concentration starts to decrease, indicating more and more VO2+
ions are converting
to VO+ ions by photogenerated charge carriers, i.e., holes, in the anolyte chamber.
To calculate Faradaic efficiency of the PEC, the light was blocked intentionally at
certain period of time during the test and a small amount (~10 ml) of the electrolyte in a
quartz cuvette with 1 cm path length was analyzed using a UV–vis spectrophotometer
(PerkinElmer Lambda 35) to determine the electrolyte concentration change.
550 600 650 700 750 800 850 9000.04
0.08
0.12
0.16
0.20
0.24
Absorb
ance (
a.u
)
Wavelength, nm
0 hr
35 hr
60 hr760 nm
Figure 5-8 UV-vis absorbance spectra of 0.01 M vanadium(IV) ions in 3 M MSA
electrolyte at different period of time of photocharging.
According to Beer-Lambert law, i.e.,
85
𝐴 = 𝜀𝑙𝑐.........................................5.4
Where:
ɛ (L· mol-1
· cm-1
): molar absorptivity of the sample measured
l (cm): the path length of the cuvette in which the sample is contained
c (mol/L-1
): the concentration of the compound in solution
The Faradaic efficiency ( F ), defined as the following equation
𝜂𝐹 =𝐹∙∆𝑛
∆𝑄 ………………………………5.5
Wherein:
Δn: the amount of reacted VO2+
during the photocharging process
F: the Faraday's constant as 96485 C/mol
ΔQ: the charge transferred during the photocharging process
is then calculated to be 84.8%.
To measure the incident photon-to-current conversion efficiency (IPCE) of the
cell, the wavelength of the incident light was controlled by a monochromator
(Optometrics) from 200 nm to 600 nm, then the IPCE was calculated according to the
following equation:[44]
𝐼𝑃𝐶𝐸 =1240 ∙𝐼𝑝
𝜆∙𝐽 𝑙𝑖𝑔 𝑡 ……………………………..5.6
Wherein:
Iph (A/cm2): the measured photocurrent density at a specific wavelength
λ (nm): the wavelength of incident light
Jlight (W/cm2): the light irradiance determined by a photodetector (Newport, USA)
86
350 400 450 500 550 600
0
10
20
30
40
50
60
IPC
E (
%)
Wavelength (nm)
3 M H2SO
4
0.01 M V in 3 M H2SO
4
3 M MSA
0.01 M V in 3 M MSA
Figure 5-9 IPCE of the PEC using a TiO2 photoelectrode in various electrolytes.
The incident photon-to-current efficiency (IPCE) of the PEC storage cell using
various electrolytes is shown in Figure 5-9. The results are in agreement with our
previous work[44] and the findings observed in chapter 5. All curves show a maximum
efficiency at 350 nm regardless of the electrolyte, which indicates the large bandgap of
TiO2 that only absorbs light in UV region. Pure H2SO4 electrolyte only gives a low IPCE
value as 2.45% due to slow reaction kinetics of water splitting reaction whereas pure
MSA electrolyte improves cell IPCE more than 7 times compared to the former. On the
other hand, vanadium redox as expected, plays a significant role by boosting IPCE of the
cell, especially with MSA electrolyte. With vanadium redox in the electrolyte, IPCE of the
cell is doubled for sulfuric acid. When sulfuric acid is replaced with MSA as the
supporting electrolyte, the highest value (45.6%) is achieved with the assistance of
vanadium redox, improving IPCE of the cell by a factor of 18.6, 9.7, and 2.5 compared to
87
pure H2SO4 acid, 0.01 M V in 3 M H2SO4 and pure MSA electrolyte respectively. Such
huge enhancement on IPCE of the cell is believed due to synergy between fast vanadium
redox kinetics and prolonged electron life time of MSA.
5.7 Conclusion
In this chapter, we investigated the electrochemical and photoelectrochemical
performance, ionic conductivity, chemical stability, reaction kinetics of vanadium redox
species in MSA electrolyte by CV, ZRA and EIS.
CV results show that MSA is chemically stable within large potential window
tested from -0.5 to 2.1 V under either dark or AM 1.5 illumination. The photocurrent is
extensively boosted in MSA-based electrolyte compared to H2SO4-based electrolyte
according to both CV and ZRA measurements, especially in the presence of vanadium
redox species.
Although the ionic conductivity of MSA is found closely comparable to that of
conventional strong acid as H2SO4 under same concentration regardless of the
involvement of vanadium redox species however, EIS Nyquist plots reveal that MSA is
capable of diminishing charge transfer resistance and interfacial capacitance at
photoelectrode/electrolyte interface under AM 1.5 illumination, especially when vanadium
redox species participates in the reaction. Other than that, EIS Bode plots manifest that
much higher electron lifetime is realized on MSA-based electrolytes in
photoelectrochemical reactions compare to H2SO4-based electrolytes. In addition, the
fast reaction kinetics of vanadium redox species shortens electron lifetime of the TiO2
photoelectrode in both acids greatly due to their quick charge scavenging ability.
The long-term (60 hrs) chemical stability measurement of all-vanadium PEC
reveals rather stable photoelectrochemical behavior of the system and XRD
88
characterization conducted on TiO2 photoelectrode shows no crystal structural change of
the material.
The cell properties, such as Faradaic efficiency and IPCE were investigated at
last. The vanadium redox species concentration is discovered to reduce after long-term
photocharging experiment and the Faradaic efficiency of the PEC is calculated to be
84.8%. IPCE of the PEC were collected using the TiO2 photoelectrode in various
electrolytes and the results match well with our previous CV, ZRA and EIS findings in
chapter 5. The highest IPCE of the PEC 45.6%, was achieved on 0.01 M V-3 M MSA
electrolyte and is attributed to the synergistic effect of fast reaction kinetics of vanadium
redox and prolonged electron life time of MSA.
89
Chapter 6
Conclusion
Photoelectrochemical cell (PEC) is an electrochemical device that provides a
practical means of transforming solar energy into electricity or chemical fuels such as H2
upon solar illumination. However, low solar-to-hydrogen conversion efficiency, lack of
cost-effective approaches to continuously produce H2 fuel regardless of the intermittent
nature of the sun slight and the very low energy density of hydrogen gas at ambient
conditions serve as a few greatest hurdles in advancing this promising technique in last
four decades.
In this study, an all-vanadium PEC was proposed and developed, and its
electrochemical and photoelectrochemical performance was investigated using common
photoelectrodes such as TiO2 and WO3/TiO2 hybrid electrodes via multiple instrumental
and electrochemical methods.
The proof-of-the-concept experiment conducted on TiO2 electrode clearly reveals
a significant photoresponse enhancement compared to photolysis of water and even
higher photocurrent on the WO3/TiO2 hybrid electrode in contact with one vanadium
redox pair. Further photoelectrochemical study using the WO3/TiO2 hybrid electrode
shows more remarkably enhanced photocurrent when two pairs of vanadium redox
couples were employed in an all-vanadium photoelectrochemical cell.
It is believed that the vanadium redox pairs serve two purposes in the
electrochemical and/or photoelectrochemical reactions: (i) quickly scavenge the charge
carriers, mitigate their recombination and thus boost the photocurrent, and (ii) shift the
dynamic equilibrium between charge carrier recombination and redox reaction toward the
latter due to fast kinetics of these vanadium redox couples.
90
In addition, a photocharge/discharge process was discovered on WO3-based
electrodes due to the formation/decomposition of hydrogen tungsten bronze (HxWO3)
under irradiation in highly acidic environment regardless of vanadium redox reactions. CV
and ZRA results reveal greatly mitigated photocurrent in pure acid electrolyte but
enhanced photoresponse of the hybrid electrode in all-vanadium electrolytes. Without
applying the bias, hydrogen tungsten bronze was formed in both pure acid and all-
vanadium electrolytes, and this finding was corroborated by EIS results, material
characterization results and proposed mechanism model. In pure acid, the formed
hydrogen tungsten bronze acts as the hurdle to prevent electron transport and thus
mitigates the photocurrent; whereas the all-vanadium electrolytes are beneficial to
revitalize photocurrent of the hybrid electrodes under illumination and render energy
storage ability under dark. In light of such a synergy effect, the alliance of vanadium
redox reaction and WO3/TiO2 hybrid electrode would potentially enable a unique dual
solar/electron storage system albeit the intermittent nature of sunlight.
To optimize the cell, the existing supporting electrolyte sulfuric acid was replaced
with methanesulfonic acid (MSA) to realize better photocatalytic property of the all-
vanadium PEC. CV, ZRA and conductivity measurements show that MSA is a chemically
stable supporting electrolyte with conductivity comparable to conventional sulfuric acid
but rendering much higher photoelectrochemical performance and therefore higher
efficiency. According to EIS results, such ability of MSA compare to H2SO4-based
electrolytes, is rooted from its prolonged electron lifetime and being capable of
diminishing charge transfer resistance and interfacial capacitance at
photoelectrode/electrolyte interface under AM 1.5 illumination, especially when vanadium
redox species participates in the reaction.
91
Although electron lifetime of the TiO2 photoelectrode is shortened greatly in both
acids with the involvement of vanadium redox species, the fast reaction kinetics of
vanadium redox species plays the key role in achieving remarkable photoresponse and
cell efficiencies due to their quick charge scavenging ability.
92
Appendix
Achievements
93
During my PhD study in Department of Materials Science and Engineering at
University of Texas at Arlington, all achievements were made and reflected in the
following public exposures, journals, patents and conferences.
PUBLIC EXPOSURE (examples):
UTA homepage UTA Shorthorn Science Daily Science Newsline
News codex American Laboratory Kurzweil News
JOURNALS:
1. Dong Liu, Zi Wei et al, Reversible Electron Storage in an All-Vanadium
Photoelectrochemical Storage Cell: Synergy between Vanadium Redox
and Hybrid Photocatalyst, 2015, ACS Catalysis, 5, 2632-2639
2. Dong Liu, Zi Wei et al, Efficient Solar Energy Storage Using A TiO2/WO3
Tandem Photoelectrode in An All-vanadium Photoelectrochemical Cell,
Electrochimica Acta, 2014, 136, 435-441
3. Dong Liu, Fuqiang Liu et al, Effect of Vanadium Redox Species on
Photoelectrochemical Behavior of TiO2 and TiO2/WO3 Photo-electrodes,
Journal of Power Sources, 2012, 213, 78-82
4. Dong Liu, Zi Wei et al, Ultra-long Electron Lifetime Induced Efficient
Solar Energy Storage by an All-Vanadium Photoelectrochemical Storage
Cell Using Methanesulfonic Acid, Journal of Materials Chemistry A,
under review
5. Zi Wei, Dong Liu et al, All-vanadium Redox Photoelectrochemical Cell:
An Approach to Store Solar Energy, Electrochemistry
Communications, 2014, 45, 79-82
94
6. Zi Wei, Dong Liu et al, Carbon Coated TiO2 Photoanode for All-
vanadium Redox Photoelectrochemical Cell as Efficient Solar Energy
Storage Device, ECS Transaction, 2015, In press
7. Zi Wei, Yi Shen, Dong Liu et al, Improved Solar Energy Storage
Efficiency of All-Vanadium Photoelectrochemical Cell using Geometry-
Enhanced TiO2 Nanobelts, ACS Nano, under review
8. Syed Dawar Sajjad, Dong Liu et al, Guanidinium Based Blend Anion
Exchange Membranes for Direct Methanol Alkaline Fuel Cells
(DMAFCs), Journal of Power Sources, 2015, In press
9. Chiajen Hsu, Shen Yi, Zi Wei, Dong Liu et al, Anatase TiO2 Nanobelts
with Plasmonic Au Decoration Exhibiting Efficient Charge Separation and
Enhanced Activity, Journal of Alloys and Compounds, 2014, 613, 117-
121
10. Noor Siddique, Amir Salehi, Zi Wei, Dong Liu et al., Length-scale
Dependent Phase Transformation of LiFePO4: An In situ and Operando
Study Using Micro Raman Spectroscopy and XRD, ChemPhysChem,
2015, accepted, DOI: 10.1002/cphc.201500299R1
PATENTS:
1. Fuqiang Liu (advisor), Dong Liu, Wei Zi, Yi Shen, A Continuous Flow
Reactor for Highly Efficient Regenerative Solar Energy Storage using
Vanadium Redox, U.S. Patent, 2015, appl. No: 62/162,976
CONFERENCES:
1. Dong Liu, Fuqiang Liu, Regenerative Solar Energy Storage via
Photocatalytic Redox Reaction, The 221st
ECS Conference, 2012,
Seattle, WA
95
2. Dong Liu, Fuqiang Liu, Photo-assisted energy storage via Vanadium(IV)
Redox species, ASM Symposium of North Texas Inter-University,
2012, Denton, TX
3. Zi Wei, Dong Liu et al, An Efficient Solar Energy Storage System: All
Vanadium Redox Photoelectrochemical Cell, The 227th
ECS
Conference, 2015, Chicago, IL
4. Zi Wei, Dong Liu, Yi Shen, et al., Enhanced Photoactivity Using TiO2
Nanobelts in An All Vanadium Redox Photoelectrochemical Cell, The
50th
anniversary TSM meeting, 2015, Austin, TX
5. Zi Wei, Dong Liu et al., All-vanadium Redox Photo-electrochemical Cell:
An New Approach to Store Solar Energy, The 49th
anniversary TSM
meeting, 2014, Arlington, TX
96
References
[1] N.S. Lewis, Toward Cost-Effective Solar Energy Use, Science, 315 (2007) 798.
[2] M. Wolf, Solar Energy Utilization by Physical Methods, Science, 184 (1974) 382.
[3] G.C. Nathan S. Lewis, Arthur J. Nozik, Michael R. Wasielewski, Paul Alivisatos, Basic
research needs for solar energy ultilization, 2005, pp. 276.
[4] N.S. Lewis, Artificial photosynthesis, American Scientist, 83 (1995) 534.
[5] M. Gratzel, Photoelectrochemical cells, Nature, 414 (2001) 338.
[6] R.N. Pandey, K.S.C. Babu, O.N. Srivastava, High conversion efficiency
photoelectrochemical solar cells, Progress in Surface Science, 52 (1996) 125.
[7] A. Fujishima, K. Honda, Electrochemical Photolysis of Water at a Semiconductor
Electrode, Nature, 238 (1972) 37.
[8] A.J. Bard, Photoelectrochemistry, Science, 207 (1980) 139.
[9] A. Heller, Conversion of sunlight into electrical power and photoassisted electrolysis of
water in photoelectrochemical cells, Accounts of Chemical Research, 14 (1981)
154.
[10] Y.V. Pleskov, Solar energy conversion in photoelectrochemical cells with
semiconductor electrodes, Progress in Surface Science, 15 (1984) 401.
[11] K. Rajeshwar, P. Singh, J. DuBow, Energy conversion in photoelectrochemical
systems — a review, Electrochimica Acta, 23 (1978) 1117.
[12] M.G. Walter, E.L. Warren, J.R. McKone, S.W. Boettcher, Q. Mi, E.A. Santori, N.S.
Lewis, Solar Water Splitting Cells, Chemical Reviews, 110 (2010) 6446.
[13] O.V. Craig Grimes, Sudhir Ranjan, Light, Water, Hydrogen-The Solar Generation of
Hydrogen by Water Photoelectrolysis, Springer US2008.
[14] S. Licht, G. Hodes, R. Tenne, J. Manassen, A light-variation insensitive high
efficiency solar cell, Nature, 326 (1987) 863.
97
[15] S. Licht, Photoelectrochemical Storage Cell, Photoconversion of Solar Energy,
World Scientific Publishing Co, 2008.
[16] G. Hodes, D. Cohen, J. Manassen, Photoelectrochemical cells, Chemtech, 11
(1981) 112.
[17] K. Fujihara, T. Ohno, M. Matsumura, Splitting of water by electrochemical
combination of two photocatalytic reactions on TiO2 particles, Journal of the
Chemical Society, Faraday Transactions, 94 (1998) 3705.
[18] S. Licht, Photoelectrochemical Solar Energy Storage Cells, Encyclopedia of
Electrochemistry, Wiley-VCH Verlag GmbH & Co, 2007.
[19] C.-C. Lo, C.-W. Huang, C.-H. Liao, J.C.S. Wu, Novel twin reactor for separate
evolution of hydrogen and oxygen in photocatalytic water splitting, International
Journal of Hydrogen Energy, 35 (2010) 1523.
[20] T. Ohno, D. Haga, K. Fujihara, K. Kaizaki, M. Matsumura, Unique Effects of Iron(III)
Ions on Photocatalytic and Photoelectrochemical Properties of Titanium Dioxide,
The Journal of Physical Chemistry B, 101 (1997) 6415.
[21] S.-C. Yu, C.-W. Huang, C.-H. Liao, J.C.S. Wu, S.-T. Chang, K.-H. Chen, A novel
membrane reactor for separating hydrogen and oxygen in photocatalytic water
splitting, Journal of Membrane Science, 382 (2011) 291.
[22] T. Simon, N. Bouchonville, M.J. Berr, A. Vaneski, A. Adrović, D. Volbers, R.
Wyrwich, M. Döblinger, A.S. Susha, A.L. Rogach, F. Jäckel, J.K. Stolarczyk, J.
Feldmann, Redox shuttle mechanism enhances photocatalytic H2 generation on
Ni-decorated CdS nanorods, Nat Mater, 13 (2014) 1013.
[23] N.A. Ramos-Delgado, M.A. Gracia-Pinilla, L. Maya-Treviño, L. Hinojosa-Reyes, J.L.
Guzman-Mar, A. Hernández-Ramírez, Solar photocatalytic activity of TiO2
98
modified with WO3 on the degradation of an organophosphorus pesticide, Journal
of Hazardous Materials, 263, Part 1 (2013) 36.
[24] I. Shiyanovskaya, M. Hepel, Bicomponent WO3 /TiO2 Films as Photoelectrodes,
Journal of The Electrochemical Society, 146 (1999) 243.
[25] N.R. de Tacconi, C.R. Chenthamarakshan, G. Yogeeswaran, A. Watcharenwong,
R.S. de Zoysa, N.A. Basit, K. Rajeshwar, Nanoporous TiO2 and WO3 Films by
Anodization of Titanium and Tungsten Substrates: Influence of Process
Variables on Morphology and Photoelectrochemical Response, The Journal of
Physical Chemistry B, 110 (2006) 25347.
[26] S. Higashimoto, T. Shishido, Y. Ohno, M. Azuma, M. Takahashi, M. Anpo,
Photocharge-Discharge Behaviors of Hybrid WO3/TiO2 Film Electrodes:
Conversion, Storage of Electrons, and the Effect of the WO3 Structure on
Rechargeability, Journal of The Electrochemical Society, 154 (2007) F48.
[27] A.J. Bard, Photoelectrochemistry and heterogeneous photo-catalysis at
semiconductors, Journal of Photochemistry, 10 (1979) 59.
[28] N. Getoff, Photoelectrochemical and photocatalytic methods of hydrogen production:
A short review, International Journal of Hydrogen Energy, 15 (1990) 407.
[29] Y.Y.G. Yu. V. Pleskov, Semiconductor Photoelectrochemistry, Springer, 2012.
[30] R.P. Allen J. Bard, Joseph Jordan, Standard Potentials in Aqueous Solution, CRC
Press, 1985.
[31] M. Gattrell, J. Park, B. MacDougall, J. Apte, S. McCarthy, C.W. Wu, Study of the
Mechanism of the Vanadium 4+/5+ Redox Reaction in Acidic Solutions, Journal
of The Electrochemical Society, 151 (2004) A123.
[32] G. Oriji, Y. Katayama, T. Miura, Investigation on V(IV)/V(V) species in a vanadium
redox flow battery, Electrochimica Acta, 49 (2004) 3091.
99
[33] G. Oriji, Y. Katayama, T. Miura, Investigations on V(IV)/V(V) and V(II)/V(III) redox
reactions by various electrochemical methods, Journal of Power Sources, 139
(2005) 321.
[34] L.G. Hepler, J.O. Hill, I.G. Worsley, Thermochemistry and oxidation potentials of
vanadium, niobium, and tantalum, Chemical Reviews, 71 (1971) 127.
[35] X. Liu, F. Wang, Q. Wang, Nanostructure-based WO3 photoanodes for
photoelectrochemical water splitting, Physical Chemistry Chemical Physics, 14
(2012) 7894.
[36] D. Liu, F. Liu, J. Liu, Effect of vanadium redox species on photoelectrochemical
behavior of TiO2 and TiO2/WO3 photo-electrodes, Journal of Power Sources, 213
(2012) 78.
[37] S. Higashimoto, N. Kitahata, K. Mori, M. Azuma, Photo-electrochemical properties of
amorphous WO3 supported on TiO2 hybrid catalysts, Catalysis Letters, 101
(2005) 49.
[38] S. Higashimoto, Y. Ushiroda, M. Azuma, Electrochemically Assisted Photocatalysis
of Hybrid WO3/TiO2 Films: Effect of the WO3 Structures on Charge Separation
Behavior, Topics in Catalysis, 47 (2008) 148.
[39] T. Tatsuma, S. Saitoh, Y. Ohko, A. Fujishima, TiO2−WO3 Photoelectrochemical
Anticorrosion System with an Energy Storage Ability, Chemistry of Materials, 13
(2001) 2838.
[40] C. Ng, A. Iwase, Y.H. Ng, R. Amal, Understanding Self-Photorechargeability of WO3
for H2 Generation without Light Illumination, ChemSusChem, 6 (2013) 291.
[41] C. Ng, Y.H. Ng, A. Iwase, R. Amal, Visible light-induced charge storage, on-demand
release and self-photorechargeability of WO3 film, Phys Chem Chem Phys, 13
(2011) 13421.
100
[42] P. Ngaotrakanwiwat, T. Tatsuma, S. Saitoh, Y. Ohko, A. Fujishima, Charge-
discharge behavior of TiO2-WO3 photocatalysis systems with energy storage
ability, Physical Chemistry Chemical Physics, 5 (2003) 3234.
[43] D. Liu, Z. Wei, C.-j. Hsu, Y. Shen, F. Liu, Efficient Solar Energy Storage Using A
TiO2/WO3 Tandem Photoelectrode in An All-vanadium Photoelectrochemical
Cell, Electrochimica Acta, 136 (2014) 435.
[44] Z. Wei, D. Liu, C. Hsu, F. Liu, All-vanadium redox photoelectrochemical cell: An
approach to store solar energy, Electrochemistry Communications, 45 (2014) 79.
[45] P.G. Dickens, M.S. Whittingham, The tungsten bronzes and related compounds,
Quarterly Reviews, Chemical Society, 22 (1968) 30.
[46] D. Yoon, A. Manthiram, Hydrogen tungsten bronze as a decoking agent for long-life,
natural gas-fueled solid oxide fuel cells, Energy & Environmental Science, 7
(2014) 3069.
[47] M.S. Whittingham, Hydrogen motion in oxides: from insulators to bronzes, Solid
State Ionics, 168 (2004) 255.
[48] P.J. Wiseman, P.G. Dickens, The crystal structure of cubic hydrogen tungsten
bronze, Journal of Solid State Chemistry, 6 (1973) 374.
[49] L. Wang, L. Yuan, X. Wu, J. Wu, C. Hou, S. Feng, Electrochromic response of
pulsed laser deposition prepared WO3-TiO2 composite film, RSC Advances, 4
(2014) 47670.
[50] I. Paramasivam, Y.-C. Nah, C. Das, N.K. Shrestha, P. Schmuki, WO3/TiO2
Nanotubes with Strongly Enhanced Photocatalytic Activity, Chemistry – A
European Journal, 16 (2010) 8993.
101
[51] Y.R. Ahn, C.R. Park, S.M. Jo, D.Y. Kim, Enhanced charge-discharge characteristics
of RuO2 supercapacitors on heat-treated TiO2 nanorods, Applied Physics Letters,
90 (2007) 122106.
[52] I. Abayev, A. Zaban, F. Fabregat-Santiago, J. Bisquert, Electronic conductivity in
nanostructured TiO2 films permeated with electrolyte, physica status solidi (a),
196 (2003) R4.
[53] L.R.F. Allen J. Bard, Electrochemical Methods: Fundamentals and Applications,
Wiley2000.
[54] G. Hodes, D. Cahen, J. Manassen, Tungsten trioxide as a photoanode for a
photoelectrochemical cell (PEC), Nature, 260 (1976) 312.
[55] L. Vayssieres, On Solar Hydrogen and Nanotechnology, Wiley2010.
[56] O. Khaselev, J.A. Turner, A Monolithic Photovoltaic-Photoelectrochemical Device for
Hydrogen Production via Water Splitting, Science, 280 (1998) 425.
[57] P.P. González-Borrero, F. Sato, A.N. Medina, M.L. Baesso, A.C. Bento, G.
Baldissera, C. Persson, G.A. Niklasson, C.G. Granqvist, A. Ferreira da Silva,
Optical band-gap determination of nanostructured WO3 film, Applied Physics
Letters, 96 (2010) 061909.
[58] L. Yang, H. Zhou, T. Fan, D. Zhang, Semiconductor photocatalysts for water
oxidation: current status and challenges, Physical Chemistry Chemical Physics,
16 (2014) 6810.
[59] X. Lu, S. Xie, H. Yang, Y. Tong, H. Ji, Photoelectrochemical hydrogen production
from biomass derivatives and water, Chemical Society Reviews, 43 (2014) 7581.
[60] C. Tang, D. Zhou, Methanesulfonic acid solution as supporting electrolyte for zinc-
vanadium redox battery, Electrochimica Acta, 65 (2012) 179.
102
[61] P.K. Leung, M.R. Mohamed, A.A. Shah, Q. Xu, M.B. Conde-Duran, A mixed acid
based vanadium–cerium redox flow battery with a zero-gap serpentine
architecture, Journal of Power Sources, 274 (2015) 651.
[62] N.-F.W. Sui Peng, Xiao-Juan Wu, Su-Qin Liu, Dong Fang, You-Nian Liu, Ke-Long
Huang, Vanadium Species in CH3SO3H and H2SO4 Mixed Acid as the Supporting
Electrolyte for Vanadium Redox Flow Battery, International Journal of
Electrochemical Science, 7 (2012) 643.
[63] Z. He, Z. Li, Z. Zhou, F. Tu, Y. Jiang, C. Pan, S. Liu, Improved performance of
vanadium redox battery using methylsulfonic acid solution as supporting
electrolyte, Journal of Renewable and Sustainable Energy, 5 (2013) 023130.
[64] G. Wang, J. Chen, X. Wang, J. Tian, H. Kang, X. Zhu, Y. Zhang, X. Liu, R. Wang,
Study on stabilities and electrochemical behavior of V(V) electrolyte with acid
additives for vanadium redox flow battery, Journal of Energy Chemistry, 23
(2014) 73.
[65] R. Solarska, R. Jurczakowski, J. Augustynski, A highly stable, efficient visible-light
driven water photoelectrolysis system using a nanocrystalline WO3 photoanode
and a methane sulfonic acid electrolyte, Nanoscale, 4 (2012) 1553.
[66] M. Zukalova, M. Bousa, Z. Bastl, I. Jirka, L. Kavan, Electrochemical Doping of
Compact TiO2 Thin Layers, The Journal of Physical Chemistry C, 118 (2014)
25970.
[67] T. Berger, T. Lana-Villarreal, D. Monllor-Satoca, R. Gómez, Charge transfer
reductive doping of nanostructured TiO2 thin films as a way to improve their
photoelectrocatalytic performance, Electrochemistry Communications, 8 (2006)
1713.
103
[68] H. Pelouchova, P. Janda, J. Weber, L. Kavan, Charge transfer reductive doping of
single crystal TiO2 anatase, Journal of Electroanalytical Chemistry, 566 (2004)
73.
[69] M. D. Gernon, M. Wu, T. Buszta, P. Janney, Environmental benefits of
methanesulfonic acid . Comparative properties and advantages, Green
Chemistry, 1 (1999) 127.
[70] D. Liu, W. Zi, S.D. Sajjad, C. Hsu, Y. Shen, M. Wei, F. Liu, Reversible Electron
Storage in an All-Vanadium Photoelectrochemical Storage Cell: Synergy
between Vanadium Redox and Hybrid Photocatalyst, ACS Catalysis, 5 (2015)
2632.
[71] M. Vijayakumar, S.D. Burton, C. Huang, L. Li, Z. Yang, G.L. Graff, J. Liu, J. Hu, M.
Skyllas-Kazacos, Nuclear magnetic resonance studies on vanadium(IV)
electrolyte solutions for vanadium redox flow battery, Journal of Power Sources,
195 (2010) 7709.
[72] Z. Tang, D.S. Aaron, A.B. Papandrew, T.A. Zawodzinski, Monitoring the State of
Charge of Operating Vanadium Redox Flow Batteries, ECS Transactions, 41
(2012) 1.
[73] X. Gao, R.P. Lynch, M.J. Leahy, D.N. Buckley, Spectroscopic Study of Vanadium
Electrolytes in Vanadium Redox Flow Battery (VRFB), ECS Transactions, 45
(2013) 25.
[74] R.P. Brooker, C.J. Bell, L.J. Bonville, H.R. Kunz, J.M. Fenton, Determining Vanadium
Concentrations Using the UV-Vis Response Method, Journal of The
Electrochemical Society, 162 (2015) A608.
[75] M. Skyllas-Kazacos, M. Kazacos, State of charge monitoring methods for vanadium
redox flow battery control, Journal of Power Sources, 196 (2011) 8822.
104
Biographical Information
Before coming to United States to pursue a higher education, Dong Liu was born
in China and lived most of his life there. He was conferred his Bachelor of Science (BS)
degree in Applied Chemistry (Materials Science) from Zhejiang University of Technology
in 2004 and his Master of Science (MS) degree in Materials Science and Engineering
from East China University of Science and Technology and Auburn University in 2007
and 2010 respectively.
In 2010 he joined The University of Texas at Arlington to pursue his doctoral
degree and worked under supervision of Dr. Fuqiang Liu in Department of Materials
Science and Engineering and received his Ph.D. degree in 2015. His primary research
area during his doctoral study was science, engineering and technology of all-vanadium
photoelectrochemical cell, while he also worked on vanadium redox flow battery and
lithium-ion battery. His academic achievements within his PhD program were reported in
Appendix of this thesis.
He is going to work in either academia or industry to keep strengthening his
professional career and plans to become a leader in both technical and managerial roles
in next 5 years after graduation.